WO2005112755A2 - Morphometric analysis of brain structures - Google Patents

Abstract

Morphometric information about brain structures can be used as an indicator of resistance/susceptibility to a disorder or abnormal behavior, therapeutic efficacy, and the presence and staging of active illness (e.g., an active disorder or abnormal behavior). For example, topological analysis of iso-surfaces from the amygdala can be used to identify cocaine-dependency.

Description

MORPHOMETRIC ANALYSIS OF BRAIN STRUCTURES

CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Patent Application Serial Nos. 60/573,138, filed on 21 May 2004 and 60/598,503, filed on 2 August 2004, the contents of which are hereby incorporated by reference in their entireties.

BACKGROUND Neuroimaging has been used to detect abnormalities in individuals that suffer from neuropsychiatric disorders. Many conventional methods for evaluating neuropsychiatric disorders rely on outwards signs or "exophenotypes" of illness. The American Psychiatric Association's Diagnostic Statistical

Manual (DSM-IV, 1994) is an example of a diagnostic procedure that uses such exophenotypes. The identification of endophenotypes, however, is expected to facilitate further characterization of neuropsychiatric behavior. Cocaine use and cocaine addiction are responsible for failure in life roles, and for enormous costs in terms of disability and direct health expenditure. Yet current treatments remain disappointing, and our understanding of risk factors for cocaine addiction and of pathogenesis remains incomplete.

SUMMARY We have discovered, inter alia, that morphometric information describing the surface topology of the amygdala can differentiate between cocaine addicts and matched control subjects. Morphometric information about the amygdala and other brain structures can also be used for comparative purposes, for example, (i) to evaluate susceptibility or resistance to a disease, disorder, or behavior; (ii) to evaluate an active disease, disorder, or behavior, or probability of contracting an active disease, disorder, or behavior; (iii) to stage a disease, disorder, or behavior; and (iv) to assess efficacy of a treatment or other therapy. In one aspect, the invention features methods of evaluating a subject. The methods include evaluating a region of the brain of the subject using a morphometric parameter and determining if a structure in the brain is altered relative to a reference morphometric parameter for a reference structure. For example, the region can include the amygdala or another brain structure listed in Table 1 (below), and the morphometric parameter can describe at least one structural feature of the amygdala or such other brain structure. Since a reduced amygdala or an amygdala with a particular symmetry property is associated with cocaine addiction, the methods can further include, if the reference represents a non-substance addicted subject and the alteration is a statistically significant reduction in the morphometric parameter relative to the reference, recommending a treatment or behavior to the subject, to avoid, prevent, or delay substance addiction, e.g., cocaine addiction or addiction to another substance described herein. Similarly, if the reference represents a substance-addicted subject and the alteration in the morphometric parameter is not statistically significant relative to the reference, the method can include recommending a treatment or behavior to the subject, to avoid, prevent, or delay substance addiction. The method can include providing a preventative or remedial therapy; e.g., for heroin addicts, the subject can be administered an oral dose of a synthetic opiate, usually methadone hydrochloride or levo-alpha-acetyl methadol (LAAM), administered at a dosage sufficient to block the effects of heroin and yield a stable, noneuphoric state free from physiological craving for opiates. With respect to cocaine addicts, see, e.g., American Psychiatric Association, Work Group on Substance Use Disorders. Practice guidelines for the treatment of patients with substance use disorders: Alcohol, cocaine, opioids. Am J Psychiatry 152(suppl):2-59, 1995. For example, the morphometric parameter can refer to a contour of a brain structure, volume of a brain structure, or symmetry-asymmetry, e.g., between left and right hemispheres. In certain embodiments, the reference amygdala is the amygdala of a normal or control subject. In other embodiments, the reference amygdala is an virtual structure whose structural features are based on the structures of the amygdalae of a plurality of subjects and a pre-selected probability value, the value representing the probability that the brain structure of one of the members of the plurality is within a constraint of the virtual brain structure. Determining whether a brain structure is altered can include comparing surface contours of the amygdala of the subject to the surface contours of the reference amygdala. Comparing can include, for example, generating a graphical output wherein an image of the surface contours of the amygdala of the subject is overlayed on an image of the surface contours of the reference amygdala. Comparing can also include evaluating an undercut or overcut at one or more positions on the surface contour. For example, the morphometric parameter is a linear measurement of over- or under-cutting, e.g., at a point on the amygdala surface, e.g., in the left or right hemisphere. In another example, the morphometric parameter can be a volumetric measurement. Further, the determining can include evaluating a volume of the amygdala of the subject and comparing the volume of the amygdala of the subject to a volume of a reference amygdala. In still another example, the morphometric parameter is a surface descriptor. In some embodiments, the methods include using MRI to evaluate a region of the brain. Evaluating a region of the brain of the subject can include voxel-based morphometry or segmentation-based morphometry. Evaluating can also include evaluating one or more surface features of the brain structure, for example, locating surface features of the brain structure and comparing one or more of the surface features to a reference surface, e.g., an isoform surface for the brain structure, e.g., an isoform surface based on the brain structure in each member of a cohort of subjects. The subject in all of the new methods can be, for example, a non-human animal, e.g., a non-human mammal, such as a non-human primate, or a human. The subject can be male or female, or adult or juvenile (e.g., pre-adolescent or adolescent). In certain embodiments, the subject is treated with a candidate therapeutic agent or a candidate therapy, prior to, during, or after imaging the subject to obtain the morphometric parameter. In other embodiments, the subject is subjected to a candidate therapy (e.g., acupuncture, surgery, psychotherapy, and conditioning). In another aspect, the invention features methods of evaluating a subject. The methods include evaluating a brain structure of the subject to obtain morphometric information about the brain structure; and comparing the morphometric information about the brain structure to corresponding reference information. In certain embodiments, the reference information is a statistical measure of the brain structure in each individual of a cohort of individuals. The brain structure can be selected from the list in Table 1 (below). The methods can include evaluating a plurality of brain structures, e.g., a set of at least two or three brain structures that interact in a circuit, e.g., the amygdala and the frontal orbital cortex. In other embodiments, the morphometric information describes a volume or surface topology of the brain structure. In some embodiments, the morphometric information describes degree of symmetry-asymmetry. In certain embodiments, each individual of a cohort is characterized by a behavioral disorder, e.g., addiction to an addictive substance, e.g., cocaine. In some embodiments, each individual of a cohort is characterized by a neuropsychiatric disorder or a neurodegenerative disorder. Examples of neuropsychiatric disorders include schizophrenia, manic depression, bipolar disorder, addictions (e.g., substance abuse, gambling, etc.), obsessive- compulsive disorder, anxiety/paranoia, autism, schizo-affective disorder; delusional disorder, psychotic disorders not elsewhere specified; antisocial personality disorder, anorexia/bulimia nervosa; and so on. Similarly socially valued traits can also be evaluated, e.g., in individuals gifted with musical talent, charm, charisma, mathematical ability, persuasion, determination, creativity, and so forth. In some embodiments, each individual of a cohort is characterized by an abnormal performance in a behavioral paradigm, e.g., a social reward paradigm, a CPT/probability paradigm, a physiological aversion/pain paradigm, a mental rotation paradigm, an emotional faces paradigm, and a monetary reward paradigm. The reference information can be obtained by methods that include imaging at least a region of the brain that includes the brain structure in each individual of a cohort; aligning images obtained from each individual, and defining an isoform surface for the brain structure that is based on a pre-selected probability value, the value representing the probability that the brain structure of one of the members of the plurality is within the constraint of the virtual brain structure. In another aspect, the invention features additional methods of evaluating a subject. The methods include evaluating a brain structure of the subject to obtain morphometric information about the brain structure; and comparing the morphometric information for a brain structure in one hemisphere relative to the corresponding brain structure in the other hemisphere to obtain a measure of the degree of symmetry-asymmetry (e.g., a symmetry index). The brain structure can be selected from the list in Table 1 (below). The methods can include evaluating a plurality of brain structures, e.g., a set of at least two or three brain structures that interact in a circuit, e.g., the amygdala and the frontal orbital cortex. The degree of symmetry-asymmetry can be used as a predictor of a disorder. For example, detection of a loss of asymmetry in the amygdala can be used to provide a recommendation or treatment for an addictive disorder, e.g., narcotic addition, e.g., cocaine addiction. The loss of asymmetry can be a loss of at least 30, 40, 50, 70, 80, or 90% of the asymmetry present in a normal or reference subject. The methods can include other features described herein. For example, undercutting of an iso-surface of the anterior and superior surfaces of amygdala, relative to a reference (such as a normal individual), can indicate that a subject requires preventive or therapeutic treatment for a narcotic dependency (e.g., cocaine dependency), as can other differences in the corticomedial and basolateral nuclei. In another aspect, the invention features methods of evaluating a candidate therapy or therapeutic agent. The methods include administering the candidate agent to a subject or providing the candidate therapy to the subject; and evaluating a brain structure, wherein a morphological change in the brain structure indicates that the candidate agent is a lead compound or composition for altering a behavioral trait, or, as may be the case, that the candidate therapy is a lead therapy. For example, the brain structure can be the amygdala or another brain structure described in Table 1. In some embodiments, the behavioral trait is substance addiction, e.g., cocaine addiction. The methods can further include evaluating the test or candidate agent in a non-human animal model for addiction. In some embodiments, the evaluating includes voxel-based morphometry or segmentation-based morphometry. Evaluating can include locating one or more surface features of the brain structure and comparing one or more of the surface features to a reference surface, e.g., an isoform surface for the brain structure, e.g., an isoform surface based on the structure in each member of a cohort of subjects. In certain embodiments, a cohort of subjects receives the candidate agent or the candidate therapy. In other embodiments, the cohort of subjects consists of subjects characterized by at least one common behavioral trait, and results of evaluating the cohort are compared to corresponding results from a cohort of control subjects that do not have the common behavioral trait. For example, the common behavioral trait can be a substance addiction, e.g., cocaine addiction. In some embodiments, evaluating includes determining over- or undercuts in the iso-surface for a first cohort of subjects relative to a corresponding iso-surface for a second cohort of subjects. In other embodiments, the iso- surface is defined by a probability function, e.g., as defined by a preselected probability value, e.g., 10, 20, 30, 40, 50, 60, 70, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99%. In another aspect, the invention features methods of evaluating a subject for one or more behavioral responses associated with substance abuse. The methods include monitoring a subject during a reward-aversion paradigm, detecting a deviation in the subjects response relative to a normal or control subject, wherein the deviation is characteristic of a substance-addicted subject; and reporting the result of the detecting. For example, reporting can recommend a preventative or remedial therapy for avoiding, preventing, or reducing substance addiction or abuse. The method can include providing a preventative or remedial therapy; e.g., for heroin addicts, the subject can be administered an oral dose of a synthetic opiate, usually methadone hydrochloride or levo-alpha- acetyl methadol (LAAM), administered at a dosage sufficient to block the effects of heroin and yield a stable, noneuphoric state free from physiological craving for opiates. In some embodiments, the deviation is characteristic of cocaine addiction. For example, the reward-aversion paradigm can be a negative- outcome based monetary-reward paradigm and the deviation can be characterized by absence of a biphasic response. In another aspect, the invention features methods of providing a population-based statistic for a brain structure. The methods include evaluating images of a brain structure, e.g., for each of a plurality of subjects; aligning the images; and determining positional information defining an virtual brain structure whose structural features are based on a pre-selected probability value, the value representing the probability that the brain structure of one of the members of the plurality is within the constraint of the virtual brain structure. For example, the positional information can represent an isoform surface. In various embodiments, the pre-selected probability value can be 10, 20, 30, 40, 50, 60, 70, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99%. The plurality of subjects can include between 5 and 5000 subjects, e.g., from 8-500, 15-50, 10-35, or 51-200 subjects. For example, each subject of the plurality can have a common characteristic, e.g., a common behavioral trait that differs from normal, a genetic marker of interest (e.g., a disease-associated marker), a common experience (e.g., traumatic stress/disaster, abuse victim, drug addiction), a common disability (e.g., a learning disability or behavioral disorder) or a common learned ability (e.g., literacy)(e.g., compare juveniles with learning v. no learning). The subjects of the plurality can have the same gender, or same age, e.g., be within a 20, 15, 10, 5, or 2 years of each other. In certain embodiments, each subject of the plurality is female, and the images are obtained at a similar phase of the menstrual cycle (e.g., the same quarter of the cycle). In some embodiments, each subject of the plurality is addicted to a substance (e.g., a narcotic, caffeine). In other embodiments, each subject of the plurality has an abnormal characteristic in a behavioral paradigm, e.g., a social reward paradigm, a CPT/probability paradigm, a physiological aversion/pain paradigm, a mental rotation paradigm, an emotional faces paradigm, or a monetary reward paradigm. In certain embodiments, the virtual brain structure is smaller than normal, e.g., in one or both hemispheres. The brain structure can be the amygdala or other structure listed in Table 1 (below). The aligning can include locating one or more of the midpoints of decussations of the anterior and posterior commissures and the midsaggital plane. It can include rotation or a nondeformation transformation. Determining positional information can include gray/white matter segmentation or evaluating signal intensity histograms. The methods can further include receiving information about the brain structure of an individual subject and comparing the received information to the information about the virtual brain structure. The methods can further include providing an estimate of risk for a behavioral trait, wherein the plurality of subjects each have a common behavioral trait. In another aspect, the invention features data-structures that include morphometric information about a brain structure (e.g., the amygdala or other brain structure). The morphometric information can be based on a statistical function dependent on a cohort of individuals with a common behavioral trait (e.g., an abnormal behavioral trait). For example, the morphometric information can include information about volume of the amygdala (e.g., a quantitative measure of volume) or information about the surface topology of the amygdala (e.g., information about the degree of undercutting or overcutting relative to a reference individual or a reference cohort or information about the surface contours, e.g., coordinates). In certain embodiments, the information about surface topology of the amygdala describes at least a part of the right amygdala, or the right and left amygdala. In other embodiments, the morphometric information describes a degree of symmetry-asymmetry for a particular brain structure (e.g., the amygdala or other brain structure). For example, a change in the symmetry index can be detected. In another aspect, the invention features databases that include a plurality of records. Each record of the plurality includes (i) information that (directly or indirectly) identifies a subject or personal or clinical information about the subject and (ii) morphometric information that includes at least one morphometric parameter or a parameter that is a function of at least one morphometric parameter. For example, the morphometric parameter can be a difference between amygdala surface position in the subject relative to an isoform surface for a cohort (e.g., a normal or affected set of individuals), a difference between amygdala volume of the subject and a reference, or a scalar value that is a function of at least two morphometric parameters. As another example, the morphometric parameter can indicate degree of symmetry- asymmetry for a particular brain structure (e.g., the amygdala or other brain structure). In some embodiments, at least one record of the database describes a subject who is addicted to an addictive substance, e.g., cocaine. In another aspect, the invention features databases that include a plurality of records. Each record of the plurality includes (i) information that (directly or indirectly) identifies a subject or personal or clinical information about the subject and (ii) a result of a comparison between morphometric information about a brain structure of the subject and corresponding information about a reference brain structure (e.g., a virtual brain structure that is a function of brain structures from a cohort of subjects or a brain structure of a reference subject, or a brain structure of the same subject, e.g., at a different time or under different conditions). The invention also features systems that include a processor, a memory, and a communications interface. The communications interface is configured to receive imaging information from an imaging apparatus, the processor is configured to process or evaluate the imagining information according to a method described herein, and store results of the processing or evaluating in the memory. The invention also provides articles of machine-readable media, having encoded instructions capable of causing a processor to effect one or more methods described herein. The hardware or software can further include a module for locating the surface of a brain structure, e.g., automatically or semi- automatically. A systems biology map as defined and described in WO 2005/020788 can include morphometric information, e.g., as described herein. Similarly, a morphometric parameter described herein can be used as one parameter in a classification method or any other method described in WO 2005/020788, e.g., to evaluate genotypes or phenotypes and correlations between such factors. An exemplary method of correlating a neuropsychiatric trait with a genetic locus can include obtaining morphometric information about one or more brain structures and genetic information from a set of individuals; generating a multi-dimensional systems biology (SB) map for each individual of the set; quantitatively sorting the individuals based on the morphometric information, e.g., using an association rule algorithm, thereby identifying a subset; and comparing polymorphisms at least one genetic locus between individuals of the subset to evaluate linkage between a polymorphism and members of the subset. For example, the comparing can include a genome scan to identify a genetic marker with a significant LOD (logarithm of the odds) score for the subset. The methods can also include comparing polymorphisms of individuals excluded from the subset, e.g., to detect whether absence of an allele is determinative. Other genetic methods (e.g., families, and linkage disequilibrium) can be incorporated. After a genetic polymorphism is associated with a trait, e.g., a trait characterized at least in part by a particular morphometric feature, a bottom up approach can be used to evaluate individuals who have the polymorphism. The individuals can be evaluated at the extremes of function, and imaged as described herein to obtain corresponding morphometric information. Typically, the individuals that are evaluated are not members of the study that linked the polymorphism to the trait. The new approaches can have at least the following applications: provide confirmatory information, pro vide. information for construction of a second model of neuropsychiatric function, and enable extrapolation of genetic information to a second population of individuals. The sorting can use criteria for at least two morphometric parameters. For example, the sorting can use a scalar or vector that is a function of the at least two morphometric parameters. In some examples, a first morphometric parameter is an indicator of an alteration in the amygdala, while the second morphometric parameter is an indicator of an alteration in the frontal orbital cortex. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The disjunctive "or" also encompasses combinations of listed alternatives. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the invention, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. In the case of conflict, the present specification, including definitions, controls. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 A, B, C, D are graphs of results of amygdala and hippocampus volumes in 27 cocaine-dependent subjects and 27 matched normal controls. FIG. 2 is a three dimensional iso-surface representation of the amygdala. FIG. 3 is a set of coronal cross-sections illustrating degree of overlap between average iso-surface representations of the amygdala for cocaine- dependent and control groups. FIG. 4 is a generalized schema for connections between the basolateral group of nuclei, corticomedial (including the central nucleus) group of nuclei, and other brain regions. FIG. 5 is a schematic of an exemplary system 300. FIG. 6 is a schematic of an exemplary computing unit. FiG. 7 is a schematic of an exemplary apparatus.

DETAILED DESCRIPTION Moφhometric information about brain structures provides useful indicators of resistance/susceptibility to a disorder or abnormal behavior, therapeutic efficacy, and the presence and staging of active illness (e.g., an active disorder or abnormal behavior). Moφhometric information includes information that describes a spatial or structural property of a brain structure or relevant part thereof. An example of a brain structure is the amygdala. In the case of cocaine addiction, the right amygdala and certain subnuclei (e.g., as mentioned below) may be particularly relevant. Other examples of brain structures are provided in Table 1. Moφhometric information can be in the form of a moφhometric parameter, e.g., a quantitative or qualitative parameter. A quantitative measure of volume is one form of a moφhometric parameter. Coordinates or equations describing a surface of a brain structure are another example. Moφhometric information can be absolute, e.g., relative to a particular coordinate-frame, or can be relative. For example, a linear distance measure of the extent of over- or under-cutting of a brain structure surface of a subject relative to a reference brain structure is a useful form of relative information. To illustrate, we have observed in one set of cocaine addicted individuals that the volume of the right amygdala is decreased about 23% and that relative to an iso-surface based on probability 0.5, addicted individuals have an undercutting in the anterior extent of about 4.5 mm. The use of moφhometric information, e.g., as described herein, will aid in the diagnosis of behavioral disorders and neuropsychiatric disorders as well as in the discovery of drugs and other therapies for treating such disorders.

Addictive Substances One common behavioral disorder is addiction to an addictive substance. There are numerous examples of addictive substances and other substances that alter mental perception (many of which are controlled substances). Examples of such substances include: amphetamines, amyl nitrite, barbiturates, benzodiazepines, butyl nitrite, caffeine, cocaine, codeine, crystal meth, designer drugs, ecstasy, DMT, glue, hallucinogens, heroin, inhalants, LSD, cannabis, marijuana, mescaline, methcathinone, ritalin (methylphenidate), nitrous oxide, opiates, PCP, psilocybin, rohypnol, sedative-hypnotics, ketamine and MDMA, steroids, tramadol/ultram, tranquilizers, valerian, vivarin, nicotine, tobacco containing products, other stimulants and depressants. Still other addictive substances include alcohol, e.g., as found in alcoholic beverages. Subjects can be evaluated, e.g., to determine a moφhometric parameter of a brain structure (e.g., the amygdala or other structure in Table 1) to assess their risk for addiction to one or more of these substances. Subjects exposed to an addictive substance, previously addicted to a substance, or presently addicted to a substance can be similarly evaluated.

Medications for the Treatment of Cocaine Addiction A number of therapeutic agents have been suggested for the treatment of cocaine addiction. These include agents that aid maintenance therapy, block euphoria, or provide withdrawal treatment. Exemplary agents that have been suggested include the following: amantadine (Supports maintenance therapy); bromocriptine (supports maintenance therapy); Buprenoφhine (blocks euphoria); Bupropion (helps achieve initial abstinence); Carbamazepine (treats withdrawal); Desipramine (treats withdrawal); Fluoxetine (treats withdrawal); Flupenthixol (treats withdrawal); Imipramine (treats withdrawal); L-DOPA (serves as replacement therapy); L-tryptophan (serves as functional antagonist); Mazindol (treats withdrawal); Methylphenidate (supports maintenance therapy; Nifedipine (blocks euphoria); and Sertraline (treats withdrawal); Diltiazem (blocks euphoria/treats cocaine cardiotoxicity); Nifedipine (treats cocaine cardiotoxicity); functional blocker monoclonal antibodies; SCH23390 (blocks euphoria; Sulpiride (blocks euphoria); Verapamil (bocks euphoria/treats cocaine cardiotoxicity). The effect of one or more these agents or any candidate therapeutic agent can be monitored by evaluating the amygdala in a subject who is being treated with such an agent. The subject can be, e.g., a human or non-human subject. Examples of candidate therapeutic agents include second-generation or other derivative forms of the above examples as well as any new candidate therapeutic agent. A similar method can be used to evaluate alternative therapies, e.g., acupuncture, traditional, and homeopathic medicines. Moφhometric information can be used to evaluate such therapies and to stage recovery, e.g., during therapy, e.g., with known, proven agents and methods or candidate agents and methods

Table 1: Exemplary Brain Regions

1. Transverse Cerebral Fissue and 36. Occipital Pole

Third Ventricle 37. Paracingulate Gyrus

2. Optic Chiasm 38. Precuneous Cortex

3. Fourth Ventricle 39. Parahippocampal Gyrus,

4. Brainstem anterior division

5. Lateral Ventricles 40. Parahippocampal Gyrus,

6. Caudate posterior division

7. Putamen 41. Parietal Operculum Cortex

8. Nucleus Accumbens 42. Postcentral Gyrus

9. Pallidum 43. Planum Polare

10. Thalamus 44. Precentral Gyrus

11. Ventral Diencephalon 45. Planum Temporale

12. Inferior Lateral Ventricles 46. Subcallosal Cortex

13. Amygdala 47. Supracalcarine Cortex

14. Hippocampus 48. Supramarginal Gyrus, anterior

15. Angular Gyrus division

16. Intracalcarine Cortex 49. Supramarginal Gyrus, posterior

17. Cingulate Gyrus, anterior division division 50. Superior Parietal Lobule

18. Cingulate Gyrus, posterior 51. Superior Temporal Gyrus, division anterior division

19. Cuneal Cortex 52. Superior Temporal Gyrus,

20. Central Opercular Cortex posterior division

21. Superior Frontal Gyrus 53. Middle Temporal Gyrus,

22. Middle Frontal Gyrus anterior division

23. Inferior Frontal Gyrus, pars 54. Middle Temporal Gyrus, opercularis posterior division

24. Inferior Frontal Gyrus, pars 55. Inferior Temporal Gyrus, triangularis anterior division

25. Frontal Medial Cortex 56. Inferior Temporal Gyrus,

26. Frontal Operculum Cortex posterior division

27. Frontal Orbital Cortex 57. Temporal Fusiform Cortex,

28. Frontal Pole anterior division

29. Heschl's Gyrus (includes HI 58. Temporal Fusiform Cortex, and H2) posterior division

30. Insular Cortex 59. Middle Temporal Gyrus,

31. Juxtapositional Lobule Cortex temporooccipital part

(formerly Supplementary Motor 60. Inferior Temporal Gyrus,

Cortex) temporooccipital part

32. Lingual Gyrus 61. Temporal Occipital Fusiform

33. Occipital Fusiform Gyrus Cortex

34. Lateral Occipital Cortex; 62. Temporal Pole inferior division

35. Lateral Occipital Cortex, superior division Endophenotvpes and Markers of Disease/Disorder Progression By evaluating moφhometric information from a plurality of subject it is possible to identify at least two types of phenotypic markers: endophenotypes and markers of disease/disorder progression (MDP). An endophenotype typically includes the following properties: (a) it provides an internal marker of a probability function for disease susceptibility or resistance; (b) it is unchanged by illness progression; and (c) it has measurable heritability / familiality. See, e.g., Almasy and Blanquero (2001) Am. J. Med. Genet. 108:42. Thus, endophenotypes may be found (but not necessarily) in unaffected siblings and parents of a subject who is affected by a disorder. Similarly, the endophenotype can be present prior to onset of the disorder. Thus, endophenotypes have high diagnostic value. An endophenotype may be defined by one or more moφhometric parameters, e.g., one or more moφhometric parameters described herein. In contrast, a marker of disease/disorder progression (MDP) is changed during the progression of a disorder. Such markers can be used to characterize the disorder, prescribe or monitor a treatment, and make other decisions (e.g., medical or financial decisions). A method for evaluating moφhometric information can include a longitudinal component that is of great value in differentiating between endophenotypes and MDPs. Such longitudinal studies include analyzing a subject at a first time and then analyzing the subject at a later time, e.g., at least one week, one, two, three, four, six, ten, or twelve months later. For example, the subject might be analyzed once a year over three to five years. In some embodiments, the subject is evaluated at approximately regular intervals. During these studies, phenotypic variables that remain unchanged, but which differ from normal (e.g., which are identified as useful for classification), are variables that can serve as endophenotypes. If the subject's outward clinical manifestations of a disorder are changing, other variables detected by evaluating neural circuit function may also change. Such variables can serve as an MDP.

Imaging Moφhometric information about structures in one or more regions of the brain can be obtained from a subject in a variety of ways. These methods typically include tomographic imaging such as MRI, PET, or CT systems, but may include any imaging method, e.g., radiological and other methods. Typically, a measuring apparatus that non-invasively obtains information about the brain (e.g., structure or function) is used. The subject to be tested is placed in an imaging apparatus and requested to lie still while images are obtained. The signals can be statistically analyzed or localized to specific anatomical and functional brain regions. The details of the processes for statistically analyzing the CNS signals and localizing the signals to specific brain regions can vary, e.g., as known to those skilled in the art. Referring now to the exemplary apparatus in FIG. 9, a magnetic resonance imaging (MRI) system 215 that may be programmed to non- invasively evaluate one or more brain structures is shown. The MRI system 215 includes a magnet 216 having gradient coils 216a and RF coils 216b disposed thereabout in a particular manner to provide a magnet system 217. In response to control signals provided from a controller processor 218, a transmitter 219 pro vides a si gnal to the RF coil 216b through an RF power amplifier 220. A gradient amplifier 221 provides a current to the gradient coils 216a also in response to signals provided by the control processor 218. For generating a uniform, steady magnetic field required for MRI, the magnet system 217 may have superconducting coils driven by a generator. The magnetic fields are generated in an examination or scanning space or region 222 in which the object to be examined is disposed. For example, if the object is a person or patient to be examined, the person or portion of the person to be examined is disposed in the region 222. The transmitter/amplifier combination 219, 220 drives the coil 216b. After activation of the transmitter coil 216b, spin resonance signals are generated in the object situated in the examination space 222, which signals are detected and collected by a receiver 223. Depending upon the measuring technique to be executed, the same coil can be used as the transmitter coil and the receiver coil or use can be made of separate coils for transmission and reception. The detected resonance signals are sampled and digitized in a Digitzer/Aray proceser 224. Digitizer/Array processor 224 converts the analog signals to a stream of digital bits, which represent the measured data and provide the bit stream to the control processor 218. A display 226 coupled to the control processor 218 is provided to display a reconstructed image. The display 226 can be a monitor, or a terminal, such as a CRT or flat panel display. A user provides scan and display operation commands and parameters to the control processor 218 through a scan interface 228 and a display operation interface 230, each of which provide means for a user to interface with and control the operating parameters of the MRI system 215 in a manner well known to those of ordinary skill in the art. The control processor 218 can be coupled to a signal processor 232 and a data store 236. The signal processor can be programmed according to a method described herein, e.g., to process raw image information. The processing can include localizing signals to a particular region of the brain.

An Exemplary Integrated System Referring to FIG. 5, an exemplary integrated system 300 can be used to produce information for a database and generate processed moφhometric information, e.g., by computing iso-surfaces or other information based on a cohort of subjects. For example, the system can include a network 305 that connects one or more imagers 350 (e.g., MRI machines)with a database server 320. The imagers 350 can deliver raw or processed information to the server 320 with information that references an individual (e.g., using an anonymous index). The database server 320 also receives similarly reference information that identifies an individual (e.g., using demographic information, an anonymous identifier, or a name) and can associate information that identifies the individual with moφhometric information and any other information, e.g., information the individual's genotype, clinical information, or information for or from a systems biology map. For example, a datastructure can be used that includes a first field with a pointer to the identifying information of the individual and a second field with a pointer to the moφhometric information for the same individual, e.g., for one or more brain structures in Table 1. The identifying information may also indicate membership of the individual in a particular cohort. In one embodiment, the system 300 also includes a statistics engine that can evaluate moφhometric information, e.g., using a method described herein. The methods and other features described herein can be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or in combinations thereof. Methods can be implemented using a computer program product tangibly embodied in a machine-readable storage device for execution by a programmable processor; and method actions can be performed by a programmable processor executing a program of instructions to perform functions of the invention by operating on input data and generating output. For example, methods can be implemented advantageously in one or more computer programs that are executable on a programmable system including at least one programmable processor coupled to receive data and instructions from, and to transmit data and instructions to, a data storage system, at least one input device, and at least one output device. Each computer program can be implemented in a high-level procedural or object oriented programming language, or in assembly or machine language if desired; and in any case, the language can be a compiled or inteφreted language. Suitable processors include, by way of example, both general and special puφose microprocessors. A processor can receive instructions and data from a read-only memory or a random access memory. Generally, a computer will include one or more mass storage devices for storing data files; such devices include magnetic disks, such as internal hard disks and removable disks; magneto-optical disks; and optical disks. Storage devices suitable for tangibly embodying computer program instructions and data include all forms of non- volatile memory, including, by way of example, semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks such as, internal hard disks and removable disks; magneto-optical disks; and CD-ROM disks. Any of the foregoing can be supplemented by, or incoφorated in, ASICs (application-specific integrated circuits). Data structures, trees, databases, and other information formats described herein can be stored in a machine accessible memory (e.g., volatile or nonvolatile memory, within a CPU or external to a CPU) or on machine-readable medium (e.g., a hard disk, CD-ROM, and so forth. An example of one such type of computer is depicted in FIG. 6, which shows a block diagram of a programmable processing system (system) 510 suitable for implementing or performing the apparatus or methods of the invention. The system 510 includes a processor 520, a random access memory (RAM) 521 , a program memory 522 (for example, a writable read-only memory (ROM) such as a flash ROM), a hard drive controller 523, and an input/output (I/O) controller 524 coupled by a processor (CPU) bus 525. The system 510 can be preprogrammed, in ROM, for example, or it can be programmed (and reprogrammed) by loading a program from another source (for example, from a floppy disk, a CD-ROM, or another computer). The hard drive controller 523 is coupled to a hard disk 530 suitable for storing executable computer programs, including programs embodying the present invention, and data including storage. The I/O controller 524 is coupled by means of an I/O bus 526 to an I/O interface 527. The I/O interface 527 receives and transmits data in analog or digital form over communication links such as a serial link, local area network, wireless link, and parallel link. One non-limiting example of an execution environment includes computers running Linux Red Hat OS, Windows XP (Microsoft), Windows NT 4.0 (Microsoft) or better or Solaris 2.6 or better (Sun Microsystems) operating systems. Browsers can be Microsoft Internet Explorer version 4.0 or greater or Netscape Navigator or Communicator version 4.0 or greater. Computers for databases and administration servers can include Windows NT 4.0 with a 400 MHz Pentium II (Intel) processor or equivalent using 256 MB memory and 9 GB SCSI drive. For example, a Solaris 2.6 Ultra 10 (400Mhz) with 256 MB memory and 9 GB SCSI drive can be used. Other environments can also be used. Many of the methods described herein can be embodied as software, e.g., as machine-executable instructions. The software can be stored on a machine- readable or accessible medium or as an article, e.g., a CD-ROM, flash memory, or in memory of an accessible server. Such methods can also be implemented on a machine. Many steps within such methods can be executed, e.g., by interaction with a user or automatically. Methods can also be implemented across a network, e.g., an intranet or Internet. For example, the network can link a health care provider and a patient, a physician (e.g., a radiologist) and a patient, and different physicians (e.g., a radiologist and psychiatrist). Communications between members of the network can be secure, web-accessible, and can include hypertext, rotatable images, and other interactive or cartographic display techniques. Information Analysis Information about the brain is highly complex. It is useful to find informative clusters of information from the highly multi-dimensional feature space that represents brain structure and function. This feature space is not only large (N is at least on the order of 10³) but also heterogeneous. Exemplary features include volumes of different brain structures, thicknesses, anisotropy, and functional properties. It is possible to reduce the dimensionality of the feature space. For example, feature classes can be normalized and values can be expressed, e.g., using variance, e.g., degree of variance from the mean. Cluster and discriminant analysis can be used to combine two or more individual factors, e.g., factors about volume, surface area, or anisotropy for one or more brain structures or clinical, demographic and genetic factors. Methods of clustering include hierarchical clustering, Bayesian clustering, k-means clustering, self-organizing maps, or shortest path analysis. In one embodiment, each factor is loaded or weighted, e.g., by selecting a coefficient that is multiplied against each factor. The magnitude of the coefficient can depend on the significance of the factor relative to other factors. Weighted factors can be summed or otherwise combined, e.g., using a mathematical function to produce a scalar value. Information with reduced dimensionality can be particularly useful for classifying subjects, e.g., for diagnostics or gene mapping. One example of information with reduced dimensionality is a scalar function that depends on both the moφhology of the amygdala and the moφhology of the frontal orbital cortex, for example, volumetric measures or surface under- or over-cut measures. See generally, e.g., US2002-0042563, US 2002-0058867, and WO 2005/020788.

Pharmacology and Pharmacogenomics It is also possible to use the methods described herein to evaluate phenotypes (e.g., by imaging) of a subject undergoing a treatment. Differences in phenotype can be detected by classification (e.g., classification trees). Then associations with a particular genotype can be detected. Other strategies can also be applied, e.g., in combination with the data analysis methods and data structures described herein. Exemplary treatments include administering an agent (e.g., a medicament) and non-invasive treatments (e.g., hyponosis, psychotherapy, etc.). Homeopathic and traditional medicines such as acupuncture as well as social behaviors can be similarly analyzed. In one embodiment, recursive partitioning can be used to evaluate results from subjects undergoing a treatment (e.g., medication or a non-invasive therapy). Classification trees can be used to determine if subjects respond differently to a treatment. In one implementation, the classification can be done blind, e.g., evaluate treated subjects and controls to detect if significant classifications objectively discriminate between treated and untreated subjects (e.g., humans and non-humans).

Virtual Reference Structures based on a Cohort of Individuals In one implementation, a virtual reference structure is created, e.g., representing a statistical function for a brain structure among a cohort of individuals, e.g., individuals with a common characteristic. For example, the cohort can be a cohort of normal controls, a cohort of disorder affected individuals, e.g., substance affected individuals, or bipolar disorder-affected individuals. Using images taken of the brain or regions thereof, brain structures can be segmented as individual structures following standardized anatomic definitions (Seidman, et al. (2002). Arch Gen Psychiatry 59, 839-849; Caviness, et al. (1996). Cereb Cortex 6, 726-736; Makris et al. (1999) Neuroimage 9, 18- 45; Breiter et al. (1997) Neuron 19, 591-611 ). Segmentation can be performed manually, semi-automatically, or automatically. Images can be registered (aligned), e.g., to a reference brain that was separate from the cohort. A probability surface (or "isoform surface" or "iso- surface") for a particular structure for each cohort can be calculated, e.g., on a voxel-by- voxel basis with the aligned data. Iso-surfaces for a pre-selected probability value (e.g., probability 0.5) are created for each cohort separately. Three-dimensional visualization of these surfaces can be used to look for systematic differences in the topology of the brain structure between cohorts, or to evaluate a brain structure of a subject in comparison to the cohort. Paradigms We can characterize the relative contributions made by each of these subregions to discrete components of reward-aversion function in different individuals using paradigms (e.g., paradigms described in Breiter, et al. (1996) Neuron 17, 875-887 Breiter, H. C, and Rosen, B. R. (1999). Ann N Y Acad Sci 877, 523-547). Thus, we can sample, for example: (1) stimulus input and representation, (2) feature extraction necessary for assessing motivational intent in others, (3) probability functions necessary for expectancy determination , (4) expectancy vs. outcome functions, (5) valuation functions, and (6) positive vs. aversive outcomes. Exemplary paradigms that can be used are further described as follows: (1) Social reward paradigm (Aharon et al., 2001) Social stimuli will consist of two sets of 40 non-famous human faces

[digitized at 600 dpi in 8-bit grayscale, spatially down sampled, and cropped to fit in an oval "window" sized 310-350 pixels wide by 470 pixels high using Photoshop 4.0 software (Adobe Systems)]. Each set will consist of 20 male and 20 female faces. Subjects will be told that they will be exposed to a series of pictures that if not interfered with, will change every eight seconds. However, if they want a picture to disappear faster, they can alternate pressing the "z" and "x" keys, whereas if they want a picture to stay longer on the screen, they can alternate pressing the "n" and "m" keys. The dependent measures of interest will be the amount of work in units of key press that subjects exert in response to the different categories of stimuli, and their resulting viewing durations. Each pair of key presses will be set to increase or decrease the total viewing time according to the following formula: NewTotalTime = OldTotalTime + (ExtremeTime-OldTotalTime) / K, where ExtremeTime was 0 seconds for keypresses reducing the viewing time, ExtremeTime will be 14 seconds for keypresses increasing the viewing time, and K was a scaling constant set to 40. If the elapsed time for the picture suφassed the total time determined by keypressing, the picture was removed and the next trial began. A "slider" was displayed left of each picture indicating total viewing time at any moment, and changing with every keypress. Subjects will be informed that the task will last 40 minutes, and that this length is independent of their behavior during the task, as is their overall payment for participating in the experiment. (2) CPT with differential probability conditions The set of experimental conditions in this study are designed to parse out differences in vigilant attention during a serial processing continuous performance task [CPT-AX(del)], involving a simple probabilistic relationship between a cue and delayed target, versus a dual processing continuous performance task [CPT-AX(int)], with a complex probability relationship between a cue and delayed target. The conditional probability of a subsequent target, given the incidence of a cue, will be the same between tasks since the CPT-AX(del) and CPT-AX(int) tasks have the same total number of cue-target pairs, and the same total incidence of true cues plus false cues. The tasks will be different in that the determination of cue-target pairs is more effortful for the CPT-AX(del) task, due to divided processing and interference suppression needs. The effortful determination of cue-target pairs will impair probability computation and lead to diminished task performance. The two paradigm conditions will involve computer presentation of an auditory letter string, with each letter spoken at a rate of 1 per second. These paradigms will have an A-B-A-C-A-C-A-B-A design where the A condition will be a simple CPT (referred to as the "QA" sequence), and the B condition will be an effortful CPT with three letters between cue and target pairs. The B and C conditions will involve either serial processing (CPT-AX(del)) or divided/dual processing (CPT-AX(int)). The CPT-AX(del) is characterized by a lack of false cues or targets between each cue ("q") and target ("a") pair, or by any interdigitated cue-target pairs (i.e., "q"_"q"_"a"_"a"), thus allowing simple probabilistic assessment of cue to target pairing with serial association of stimulus and response. The CPT-AX(int) has false cues or targets between pairs of cues and targets, and has cue-target pairs interdigitate together so that commingled pairs were possible, thus preventing simple counting or rehearsal procedures (i.e., forcing subjects to maintain two or more counts), and increasing the effort needed for probabilistic assessment of cue to target pairing. Each B and C epoch will last 90 seconds, while the baseline A epochs will last 60 seconds. There will be a target to distracter ratio of .13 for both A and B conditions, and the number of cue-target pairs will be the same. Subjects will respond with a magnet compatible button press, so that reaction time and accuracy could be recorded. The order for performing the CPT-AX(del) and the CPT-AX(int) will be counterbalanced across subjects. See, e.g., (Breiter & Rosen, 1999; Seidman et al., 1998). (3) Physiological Aversion (Thermal Pain) Subjects will be informed in detail about the nature of the experiment, and the temporal sequence of procedures, including rating methods. These rating will involve rating on a scale from 0 (no pain) to 10 (maximum pain) their perception of the pain they experienced, after the functional run. Thermal stimuli will be delivered using a modified Peltier based thermode (Medoc, Haifa, Israel). One scan will be performed during which a base temperature of 35 °C (30 s) (condition A), a warm stimulus of 41 °C (25 s) (condition B), and a target temperature of 46°C (condition C) will be interleaved. The thermode will be set to change the temperature at a rate of 4 °C/s. Thus, it will take 2 s to reach 41 °C from the baseline and 2 s to return to baseline, while for the 35-46°C contrast, the delay will be 4 seconds. The delays were not part of the baseline (30 s) or stimulus (25 s) times. The three stimuli will be interleaved in a block design: A- B-A-C-A-C-A-B-A. (4) Mental rotation The figures will be the original Shepard and Metzler (1971) objects.

(Cohen et al., 1996) They will thus consist of three-dimensional perspective drawings of 10 cubes arranged in chiral patterns and viewed from a variety of rotation angles. Two task variants will be used. In a control condition, subjects will be shown a pair of figures, half of which are identical, and half of which are mirror-reversed shapes. Each of the 10 possible angled-shapes (0 - 180° in 20° increments) will appear in each type of pair. The stimulus ordering will consist of a set of blocks, so that each of the stimuli appear once before any stimulus appears twice, and each appears twice before any appears three times, and so forth. Within each of these blocks, the stimuli will appear in random order except that the same stimulus will not appear twice within three successive trials. Moreover, half of the pairs within each block will include identical figures and half include mirror-reversed figures. No more than three consecutive trials can have the same response. The second version of the task (a rotation variant) will be identical to the first except that the members of each pair will be presented at different orientations. The left member will always be presented so that the major axis is vertical. The right member will be presented at nine possible angles (20 - 180° in 20° increments) from vertical. Three sets of these rotation trials will be used (and 4 sets of control trials), which will include rotations around different major axes. One set will include rotations around the x-axis, another around the y-axis, and another around the z-axis. These stimuli will be presented in separate sets. Within each set, the stimulus trials will be ordered so that each orientation appears once before it appears again, once with identical stimuli and once with mirror-imaged stimuli, within each balanced subgroup of 18 trials. The same orientation will not appear twice within three consecutive trials. A third "resting" or fixation condition will be interleaved between the "control" and "rotation" tasks. Subjects will be asked to look at each pair, and to decide whether the figures are identical or are mirror-images and to indicate their choice by pressing one of two buttons. In the control condition, subjects will be asked to simply respond as quickly and accurately as possible. In the rotation condition, they will be told to visualize the right-hand stimulus rotating until it is aligned with the left-hand stimulus, and then to decide whether the two shapes are identical or are mirror reversed. (5) Emotional faces Faces used in these experiments will be from Ekman and Friesen (1976), see, e.g., Neuron 17, 875-887. They will have been standardized by (a) digitization, (b) scaling of extents, (c) normalization of contrast across all expressions for each of the individuals utilized (N=8), and across all individuals in the cohort., and (d) fitting with an oval mask to minimize the observation of hair. The experiment will employ an A-B-A-C-A-D-A-B-A-D-A-C-A-B-A design with equal length epochs of tachistoscopic-like presentations of the faces as. In A, subjects will see 180 presentations of 8 faces in random order; neutral expressions (200msec) will be followed by a fixation point (300msec). In B, C, and D, subjects will see faces with one emotion presented 180 times per epoch with the same timing parameters as in A. These facial expressions will be: happy (B), angry (C), and fearful (D). The order in which these blocks of facial expression are presented will be counterbalanced by emotion, and by epoch order within run. This will be a covert paradigm design with passive viewing of tachistoscopic-like face presentations, and use of the same 8 individuals for each expression presented in random order per epoch. (6) Monetary expectancy, gains, and losses In this experiment we seek to map the hemodynamic changes that anticipate and accompany monetary losses and gains under varying conditions of controlled expectation and counterfactual comparison. The display will consist of either a fixation point or one of 2 disks ("spinners"). Each spinner will be divided into 2 sectors. Both spinners will offer the same outcomes, a gain of $+10 or a loss of $-8, but the likelihood of the gain will be high (0.66) on the "good" spinner and low (0.33) on the "bad" spinner. The relative areas of the spinner assigned to the two outcomes represent the likelihoods. Thus, on the good spinner, 66% of the area is colored green and labeled $+10, and the remaining 33% of the area is colored red and labeled $-8; on the bad spinner, the colors and labels are reversed. Providing larger gains than losses will be implemented to compensate for the tendency of subjects to assign greater weight to a loss than to a gain of equal magnitude. Before the game begins, subjects will be shown each spinner 3 times so as to learn its composition. Each trial will consist of (1) an "expectancy phase," when a spinner is presented and an arrow spins around it, and (2) an "outcome phase" when the arrow lands on one sector and the corresponding amount is added to or subtracted from the subject's winnings. During the expectancy phase, the image of one of the 2 spinners will projected for 10 sec, and the subject will score their emotional response to the displayed spinner (or fixation point) using a potentiometer. During the outcome phase, the arrow will land on one of the sectors and flicker for 9.5 seconds, indicating how much they won or lost. During this time, subjects will score their emotional response to the observed outcome. After 9.5 seconds, a 0.5 second mask will appear. On fixation-point trials, an asterisk will appear in the center of the display for 19.5 sec, followed by the 0.5-sec mask. The pseudo-random trial sequence will be fully counter-balanced to the first order so that trials of a given type (spinner + outcome) are both preceded and followed by the same number of 4 spinner/outcome combinations and 2 times by fixation-point trials. Subjects will observe 24 trials of the +$10 outcome, 24 trials of the -$8 outcome, and 16 trials of spinner baseline. A "dummy" trial will be inserted at the beginning and end of each run for counterbalancing, allowing 18 trials per run for 4 runs. Runs will be separated by 2 min rest periods. The same trial sequence will be used for all subjects, generating winnings of $48, to which will be added the $50 endowment.

Connectivity of Brain Regions FIG. 4 illustrates a generalized schema for connections between the basolateral group of nuclei, corticomedial (including the central nucleus) group of nuclei, and other brain regions. The regions grouped on the left side of the figure represent regions with connections through the basolateral group of nuclei. The regions grouped on the right represent regions with connections through the corticomedial group of nuclei. In general, these two groupings have very distinct functional outputs, as noted in the boxes with dashed lines. Abbreviations used for anatomy reflect terminology used in this manuscript: sublenticular extended amygdala (SLEA) and nucleus accumbens septi (NAc). The arrows suggest some of the known connections between these regions. The subnuclei of the corticomedial group receive their primary afferents from the thalamus, entorhinal cortex, and ventral tegmentum. Their primary efferents are to the lateral hypothalamus, brainstem, and the ventral tegmentum. This network sets the corticomedial nuclei up to receive input about potential goal-objects or aversive events via relays through the thalamus, and to prepare the body for impending behavior through the brainstem via autonomic responses. The interaction with the lateral hypothalamus facilitates the determination of whether or not perceived goal-objects or events will fulfill or worsen potential deficit states behind incentive motivation (see generally, e.g., Breiter, H. C, and Gasic, G. P. (2004). A General Circuitry Processing Reward/ Aversion Information and Its Implications for Neuropsychiatric Illness. In The Cognitive Neurosciences, M. Gazzaniga, ed. (Cambridge, MIT Press)). Afferents from the lateral hypothalamus through the anterior nucleus of the thalamus to the cingulate cortex cross afferents and efferents from the basolateral subnuclei to the cingulate cortex. In contrast to the corticomedial group, the basolateral group of subnuclei receive primary input from unimodal and heteromodal cortices, paralimbic cortices (and thus the hippocampus via the entorhinal cortex), and ventral tegmentum. Its outputs include the paralimbic cortices, basal ganglia, and thalamus. Functionally, the connections with the nucleus accumbens, sublenticular extended amygdala (SLEA), and other reward aversion regions in paralimbic cortex allow basic assessment of goal-object features necessary for valuation and expectancy functions (reviewed in Breiter & Gasic, 2004, supra). Reciprocal connections to cingulate cortex allow the output of these reward/aversion functions to be evaluated against the internal state of the organism from the lateral hypothalamus, and focused on behavioral planning, and implementation via cingulate cortex connections to supplementary motor cortex, premotor cortex, and primary motor cortex.

EXAMPLES The invention will be further described in the following examples, which do not limit the scope of the invention described in the claims.

Example 1 Drug addiction is a chronic relapsing disorder in which compulsive drug- seeking and drug-taking behavior persist despite serious negative consequences. This study examined the volumes of the amygdala and hippocampus in cocaine addicted subjects and matched healthy controls, and determined that the amygdala, but not the hippocampus, was significantly reduced in volume.

Topological analysis of iso-surfaces from the amygdala indicated that the iso- surface of the cocaine-dependent group undercut the anterior and superior surfaces of the matched healthy controls, implicating a difference in the corticomedial and basolateral nuclei. Amygdala volume in cocaine addicted individuals did not correlate with measures of anxiety or depression, or any of the three measures of the amount of cocaine used over the short- or long-term. Together, these findings argue for a primary event early in the course of cocaine use or a condition that predisposes the individual to cocaine-dependence by affecting the amygdala. We evaluated the potential for volumetric and topological abnormalities in the amygdala of cocaine dependent subjects by using segmentation-based moφhometric analysis which provides absolute quantitative volumetric measures (Caviness, et al. (1996). Cereb Cortex 6, 726-736; Filipek et al. (1994) Cereb Cortex 4, 344-360; Kennedy et al. (1989) IEEE Transactions on Medical Imaging 8, 1-7; Makris et al. (1999) Neuroimage 9, 18-45). Given reported volume increases and decreases for other regions reported in the literature for cocaine dependent subjects, the hypothesis was stated in terms of differences and two-tailed t tests were employed throughout to test it. To act as a comparison for this a priori focus on the amygdala, another region to which the amygdala is connected (and which is involved in reward/aversion function), namely the hippocampus, was also evaluated. Structural imaging data from two cohorts of cocaine dependent subjects studied with cocaine infusions [N=27]; (Breiter et al. (1998) Paper presented at: Society for Magnetic Resonance in Medicine; Breiter et a. (1997) Neuron 19, 591-611) were combined. Subjects were recruited on the basis of: (1) a personal history of cocaine dependence that met DSM-IVR criteria, or (2) a personal history with four features: (a) no psychiatric illness, (b) no psychotropic medication, (c) no psychiatric hospitalization, and (d) no family history of psychiatric illness, medication or hospitalization. Subjects with a personal history of cocaine dependence were formally diagnosed via SCID-I and psychiatric interview. Substance use was quantified via the Addiction Severity Index (ASI), as reported previously. ASI measures of drug use included estimation of (1) total years of drug intake [9.5 + 8.4 (1 - 27)], (2) number of days in the past month they had used cocaine [16.3 + 8.5 (3 - 30)], and (3) amount of money spent on cocaine in the past week prior to neuroimaging [$302.5 ± 274.9 (20 - 1000)]. Age of drug use onset was 25.2 + 7.1 years of age (range, 12 - 36 years). Potential anxiety and depressive symptoms were assessed via the Hamilton Anxiety (HAMA) [4.08 + 3.84 (0 - 15, test range = 0 - 54)] and Hamilton Depression (HAMD) Scales [6.64 + 5.82 (0 - 23, test range = 0 - 52)]. Control subjects were accrued as a general control population for moφhometric studies of psychiatric populations (Goldstein et al. (1999) Arch Gen Psychiatry 56, 537-547; Gooldstein et al. (2002) Arch Gen Psychiatry 59, 154-164; Seidman, et al. (1999). Biol Psychiatry 46, 941-954; Seidman, et al. (2002). Arch Gen Psychiatry 59, 839-849). Subjects with a negative personal and family psychiatric history were further evaluated for latent psychopathology via the short- form of the Minnesota Multiphasic Personality Inventory [MMPI- 168] (Vincent and Castillo, 1984 J Clin Psychol 40, 400-402), and excluded if they had clinical scales above. They were further administered the substance use section of the Schedule for Affective Disorders and Schizophrenia (SADS). Healthy controls were individually matched to cocaine dependent subjects on the basis of age [COC=33.9 years (range 26-45 years); NC=35.6 years (range 23-46); (p = 0.3)], gender [COC=23/27 male/total; NC=21/27 male/total; (p = 0.5)], and handedness [COC=26/27 right-handed; NC=26/27 right-handed; (p = 1.0)]. Race [COC=17/27 Caucasian; NC=25/27 Caucasian; (p = 0.007)] and education [COC=12.5 years (range 10-16); NC=14.9 years (range 12-18); (p < 0.0001)] were not perfectly matched, and hence were used as covariates in the data analysis. Given a focus on amygdala and hippocampal volumes, both total and lateralized volumes, the Bonferroni correction for multiple comparisons was p < 0.05/6 = 0.0083 in this sample. MRI scans were acquired at the NMR Center of the Massachusetts

General Hospital (MGH) using a 1.5 Tesla General Electric Signa scanner (Milwaukee, WI), and TI -weighted 3D SPGR sequences optimized for segmentation at the MGH CMA (See generally Breiter et al. (1997) Neuron 19, 591-611; Kennedy et al., supra and Makris et al, supra). For healthy controls, 60 contiguous coronal slices (through-plane resolution = 3.1mm) were acquired and positionally normalized. For cocaine dependent subjects, 60 contiguous sagittal slices (through-plane resolution = 2.8mm) were acquired, and positionally normalized and resampled following standard CMA procedures to the spatial resolution of the control cohort. Images were processed and analyzed at the MGH Center for

Moφhometric Analysis (CMA) analyzed using Sun Microsystems, Inc. computer workstations. Images were positionally normalized by imposing a standard three-dimensional coordinate system on each three-dimensional MR scan using the midpoints of the decussations of the anterior and posterior commissures, and the midsagittal plane at the level of posterior commissure, as points of reference for rotation and (nondeformation) transformation. Positional normalization overcomes potential problems caused by variation in head position across subjects during scanning. Gray-white matter segmentation was performed on each TI -weighted, positionally normalized, 3D coronal scan using a semi-automated intensity contour mapping algorithm or cerebral exterior definition and signal intensity histogram distributions for demarcation of gray- white matter borders. With this technique, borders were defined as the midpoint between the peaks of the bimodal distribution for any given structure and its surrounding tissue. This technique yields separate compartments of overall cortex and subcortical gray structures corresponding to the natural tissue boundaries distinguished by the signal intensities in the TI -weighted images (Filipek et al., supra). For this study, the hippocampus and amygdala were segmented as individual structures, following standardized anatomic definitions (Breiter et al., 1997, supra; Caviness et al., supra; Makris et al., supra). Segmentation was performed by a technician blinded to subject diagnosis, and with a randomly ordered sequence of subjects. Twenty-seven cocaine dependent subjects and twenty-seven matched control subjects were registered (aligned) to a reference brain that was separate from either cohort. Amygdala probability for each group was calculated on a voxel-by-voxel basis with the aligned data. Iso-surfaces for the probability 0.5 were created for each group separately. Three-dimensional visualization of these surfaces was used to look for systematic differences in the topology of the amygdala between groups. All subjects were segmented by a trained technician who was "blind" to group status. Intra-rater reliability, as assessed by intraclass correlation coefficient [ICC] (Fleiss and Shrout (1977) Am J Public Health 67, 1188- 1191), was very good at 0.84 for the total amygdala volume. Volumes, calculated in cubic centimeters, for each individual structure were derived by multiplying the number of voxels assigned to that structure on each slice by the slice thickness and summing across all slices in which the structure appeared (Kennedy et al., supra). Estimates of location and scale of the moφhometric volumes were evaluated by separate repeated-measures analyses of variance. In each case, the within-subjects variables were anatomic volume, with subject group as a between-subjects variable. Covariate analysis was performed on years of education and age. To control for the effect of race and gender, a separate repeated-measures analyses of variance was performed of the majority gender and race. To rule out gross volumetric effects, total brain, cerebral cortex, and cerebral white matter volumes were compared for a randomly selected subset of cocaine addicted subjects (n = 10) and their matched controls (n = 10). No significant differences were noted for these moφhometric measures [total brain F (1,18) = 1.4, p = 0.25; total cerebral cortex F(l,18) = 0.05, p = 0.83; total cerebral white matter F(l, 18) = 1.7, p = 0.21)]. For correlation analysis, the following variables were available for the 27 cocaine subjects: Hamilton Anxiety Index, Hamilton Depression Index, years of cocaine use, days of cocaine use in previous month, money spent on cocaine in the previous month, and age of drug use onset. A Log transformation was performed on each variable to ensure approximately normally distributed data. Pearson coπelations were then calculated to test for significant relationships between each of the Log transformed variables and total amygdala volume and total hippocampus volume. The symmetry of the measured amygdala and hippocampal volumes was estimated by a symmetry index (Galaburda, 1987):

Symmetry index = 100 x Left volume - Right volume 1/2 (Left volume - Right volume)

where the Right volume and Left volume are the same anatomic region in the right and left hemisphere. Positive values indicate that the anatomic region of concern is larger in the left hemisphere. This symmetry index is a unit-less quantity, very much like a coefficient of variation. For coπelation analysis, the following variables were available for the cocaine subjects: Hamilton Rating Scale for Anxiety, Hamilton Rating Scale for Depression, years of cocaine use, days of cocaine use in previous month, money spent on cocaine in the previous month, age of drug use onset, , baseline craving from the TCACS, low-dose unblinded drug-induced craving, moderate-dose blinded drug induced craving, and Addiction Severity Index composite scores for Social and for Legal function. A Log transformation was performed on each variable to ensure approximately normally distributed data. Pearson correlations were then calculated to test for significant relationships between each of the Log transformed variables and amygdala volumes and hippocampus volumes. These volumes, for both structures, included the left and right volumes, total volume, and the symmetry index. The absolute volumes of various brain structures were compared. Boxplots of amygdala and hippocampus volumes in these 27 cocaine-dependent subjects and 27 matched normal controls are shown in FIG. 1. The dispersion measure is presented as a diamond, and the horizontal line through the data represents the whole-group mean. Blinded segmentation revealed smaller volumes in cocaine dependent subjects for right (1.34 ± 0.06 cc, mean ± SE), left (1.33 ± 0.06 cc), and total amygdala (2.67 ± 0.10 cc), relative to control subjects for right (1.75 ± 0.06 cc), left (1.52 ± 0.06 cc), and total amygdala (3.27 ± 0.10 cc) (FIG. 1A). An analysis of variance yielded a significant effect for the right (F(l,52) = 22, p < 0.0001) (FIG. IB) and total amygdala (F(l,52) = 17, p < 0.0001) volumes, and an effect for left amygdala volume that did not meet the Bonfeπoni coπection (F(l,52) = 5.3, p < 0.02) (FIG. 1C). Analysis of variance, when assessing a more homogeneous sample, continued to demonstrate a significant effect for diminished amygdala volume in cocaine dependent subjects. Specifically, in Caucasian males (N=14 addicted, 1 controls), amygdala volume remained smaller in cocaine dependent subjects for right (1.33 ± 0.08 cc, mean ± SE), left (1.35 + 0.08 cc), and total amygdala (2.68 + 0.14 cc), relative to control subjects for right (1.78 + 0.07 cc), left (1.63 + 0.07 cc), and total amygdala (3.41 + 0.12 cc). An analysis of variance produced a significant effect for the right (F(l,31) = 19, p < 0.0001), and total amygdala (F(l,31) = 16, p < 0.0004) volumes, but not the left amygdala volume which did not meet the Bonfeπoni threshold (F(l,31) = 6.9, p < 0.01). Symmetry coefficients between right and left amygdala volumes were -

0.015 ± 0.042 for cocaine dependent subjects and -0.134 ± 0.039 for healthy control subjects. The distribution of symmetry coefficients was symmetric for cocaine dependent subjects (two-tailed t(26) = -0.35, p = 0.73), but was rightward for matched healthy controls subjects (two-tailed t(26) = -3.43, p = 0.002). Asymmetry is a distribution of symmetry coefficients that is significantly non-zero. The symmetry coefficients were further different by an analysis of variance (F(l, 52) = 4.3, p < 0.044). Coπelational analysis was performed between four sets of clinical measures and four moφhometry measures (i.e., the amygdala symmetry coefficient, right amygdala volume, left amygdala volume, and total amygdala volume). Coπelational analysis of depressive and anxiety symptoms with amygdala volumes were not significant (all p > 0.05). Coπelation of amygdala volumes to (1) years of cocaine use, (2) days of use in past month, (3) amount of money spent on drug one week before scanning, and (4) age of beginning drug use were also not significant (all p > 0.05). Coπelation of amygdala volumes to (a) baseline craving ratings, (b) craving ratings secondary to unblinded low-dose cocaine infusion, and (c) craving ratings secondary to blinded moderate-dose cocaine infusion were also not significant, excepting a suggestive result for baseline craving. Namely, baseline drug craving (i.e., drive to use cocaine during the 24 hour period before experimental evaluation) produced a coπelation of R = -0.46, p < 0.04 with the amygdala symmetry coefficient. Coπelation of amygdala volumes to the Addiction Severity Index (ASI) for Social function and for Legal function were further not significant, excepting for four suggestive results. The ASI Social index produced a coπelation of R = -0.61 , p < 0.03 with left amygdala volume, and of R = -0.46, p < 0.12 with total amygdala volume. The ASI Legal index produced a suggestive coπelation of R = -0.43, p < 0.17 with left amygdala volume, and of R = -0.40, p < 0.20 with the amygdala symmetry coefficient. In contrast to the amygdala results, a similar blinded segmentation analysis did not identify significant differences in hippocampal volumes in cocaine dependent subjects and matched healthy controls. Iso-surfaces of the amygdala depicting the probability 0.5 were constructed for the group of cocaine dependent subjects (N = 27) and the group of controls (N = 27). The probability 0.5 surface of the cocaine dependent group undercuts the healthy control surface in a continuous fashion from an anterior- inferior position to a posterior-superior position (FIG. 2). This undercut is observed on both lateral and medial aspects of the amygdala. Maximum undercutting of the control group iso-surface is 4.5 mm in the anterior extent and less across the superior surface of the amygdala to its posterior extent, representing regions where healthy controls have amygdala and the cocaine dependent subjects do not. A small region where the healthy control iso-surface undercuts that of the cocaine dependent group is observed along the inferior extent of the amygdala (FIG. 2). On the left side of FIG. 2, the right and left lateral ventricles as well as the right and left amygdala are shown in three dimensions. The average amygdala iso-surface for the right amygdala in the control and cocaine- dependent subjects is included on the right side of the figure: the averaged amygdala of 27 normal controls and the average amygdala of 27 cocaine- dependent subjects are co-registered and superimposed. The amygdala of the cocaine-dependent subjects is encapsulated within the larger average amygdala of the normal controls. The figure illustrates a difference in size between the two groups in the anterior, superior and lateral amygdala regions. The subnuclei of the amygdala that are potentially affected by this pattern of difference between healthy and cocaine dependent subjects may belong to the corticomedial or the basolateral groups (Mai et al, 1997) (FIG. 3 (top, middle, bottom)). Coronal cross-sections illustrate the degree of overlap between average iso-surface representations of the amygdala for cocaine-dependent and control groups. Compared to the atlas, the detected differences could be inteφreted as volumetric decrease related to the following amygdala nuclei: lateral, basomedial and basolateral nuclei of the basolateral group and anterior cortical, medial and central nuclei of the corticomedial group. When the amygdala iso-surfaces are juxtaposed with those of the hippocampus, the posterior hippocampal junction shows minimal surface differences between the two groups. In this study, segmentation-based moφhometry found the amygdala volume of cocaine-dependent subjects to be significantly less than in matched controls (FIG. 1 A, B, C). This difference was most pronounced in the right amygdala, which was decreased in volume approximately 23% (versus an approximate 13% volume decrease for the left). This difference remained when evaluated for effects of race, years of education, and gender. Although an analysis of covariance showed a significant effect of age, it did not abrogate the observed amygdala volume differences between addicts and controls. The amygdala volumes of cocaine dependent subjects were similar for each hemisphere, whereas those of their matched controls had clear laterality differences. Amygdala volume in addicts did not coπelate with (1) measures of anxiety or depression, (2) any measure of the amount of cocaine use, or (3) age at which cocaine use began. In all of these clinical and drug use measures, excepting the Hamilton Rating Scale for Anxiety (HRSA), there was considerable variation in subject data, making a Type II eπor for the absence of coπelation less likely. Hippocampal volumes, which were set up as a potential contrast to those of the amygdala, did not significantly differ between addicts and controls (with 4.9% and 6.4% differences for the right and left hippocampus), supporting the specificity of the amygdala findings (FIG. ID). Topological analysis of amygdalas from addicts and unaffected controls illustrated that the iso-surface of the unaffected controls completely covered the iso-surface of addicts, except for the inferior most surface. This topological analysis suggests that the differences lie in nuclei of the corticomedial and basolateral groups. These data support at least three general conclusions. First, in a comparison of 27 cocaine dependent subjects and matched healthy controls, the amygdala, compared to connected regions such as the hippocampus, was significantly reduced in volume. These amygdala volume differences between the two cohorts were related to an absence of amygdala laterality in addicted individuals. These patterns of volume and laterality differences were distinct from those observed for the amygdala and hippocampus in a number of other neuropsychiatric illnesses with which addiction is commonly comorbid. Second, topological differences point to a complex set of potential changes in the corticomedial and basolateral subnuclei, or to one subnucleus near the center of the amygdala. Third, the reduction in amygdala volume was not coπelated with any measure of drug use, or symptom severity score, and its dispersion (i.e., variance) estimates were similar for both cohorts. In concert, these findings argue in favor of a primary event early in the course of cocaine use, or a condition that predisposes the individual to cocaine-dependence by affecting the amygdala. The amygdala is composed of more than a dozen sub-nuclei, which can be generally segregated into two groups, the corticomedial (including the central nucleus) and the basolateral groups (Figure 4), on the basis of cellular origins and other factors. As displayed in Figure 3, the regions of the amygdala affected by diminished volume in that structure in cocaine dependent subjects approximate regions containing subnuclei of both the corticomedial and basolateral groups. It is possible, though, that this diminished volume reflects effects from a relatively restricted set of subnuclei, distributed near the center of the amygdala, and producing a general inward retraction. Although amygdala volume differences in the cuπent study were consistent with the presence of moφhometric decrements for paralimbic regions connected to it, the amygdala measures were a multiple of the measures reported for them. These amygdala findings supplement and consistent with the observation of decreased fractional anisotropy for white matter regions connecting FOC and aCG with the amygdala (Lim et al., (2002) Biol Psychiatry 51, 890-895). The observed volumetric data is distinct from that reported for manic depressive illness and Alzheimer's, where both the amygdala and hippocampus show decreased volumes. This volumetric pattern is also distinct from those reported for major depressive disorder or post-traumatic stress disorder, both of which have been coπelated with some degree of diminished hippocampal volume. The presence of a decreased amygdala volume in all 7 individuals with 1-

2 years of cocaine use (i.e., 26% of the cohort), suggest that the amygdala volume changes represent a very early neuroadaptation. Alternatively, the decreased amygdala volume may precede cocaine use and represent a preexisting marker of illness susceptibility, or endophenotype. Since cocaine use varied substantially in our subjects (1-27 years), the observed changes in the amygdala do not depend on length of use and can indicate a developmental predisposition for addiction. In conclusion, it is possible to use a longitudinal study of individuals at risk who subsequently may or may not develop the disorder, or a study of the trait in individuals recruited in a family-based design to refine the role of a structural difference. It is also possible to use the observed changes in the amygdala as a brain-based phenotypic marker for illness which could facilitate future human genetic studies aimed at discovering the genes that modify the risk for addiction. Defects in the amygdala function can be indicate that substance addicted individuals, and particularly cocaine addicted individuals, require severe aversive consequences or similar negative conditioning in order to abstain for substance use. The observed decrease in amygdala volumes in cocaine- dependent subjects (spanning a range of years of drug use) versus matched controls indicates that this change represents an early neural indicator of cocaine use or a heritable trait that increases the risk of drug-dependence by impairing the amygdala's role in allowing an individual to foresee the negative consequences of a planned action.

Example 2 Bipolar disorder (BPD) is one of the most severe neuropsychiatric disorders at any age and among the most disabling psychiatric conditions affecting children DelBello et al (2004) Bipolar Disord., 6(l):43-52.). The following example used MRI to evaluate brain regions involved in the pathophysiology of pediatric BPD. inclusion criteria for this study were: DSM-IV diagnosis of BPD, age 6- 16 years, and right-handedness. Male and female subjects of all ethnicities were recruited. Healthy controls, all right handed, had no DSM-IV Axis I diagnosis on structured and clinical interviews, and had no family history of affective disorders or psychotic disorders in first-degree relatives. Exclusion criteria were: major sensorimotor handicaps; full-scale IQ < 70 or learning disabilities; history of claustrophobia, head trauma, loss of consciousness, autism, schizophrenia, anorexia or bulimia nervosa, alcohol or drug dependence/abuse (during 2 months prior to scan, or total past history of > 12 months), active medical or neurologic disease, metal fragments or implants; history of electroconvulsive therapy; cuπent pregnancy or lactation. Data from 63 subjects who were scanned as part of an ongoing neuroimaging study are included in this report: 44 children with DSM-IV BPD and 19 healthy controls. All children underwent diagnostic semi-structured (Kiddie Schedule for Affective and Schizophrenic Disorders: Epidemiologic Version-KSADS-E) and clinical interviews by board-certified child psychiatrists. Final DSM-IV diagnoses were established by the consensus diagnosis of clinical and structured interviews. Each child received a physical and neurological examination (including Tanner Staging (TS): a I-V scale of pubertal development), and cognitive testing. The age of onset was determined by the parent report on the structured interview. Age of onset of each individual symptom was obtained and age of onset of illness was determined as the time when a child met full threshold diagnostic criteria. Children received several subtests of the Wechsler Intelligence Scale Three (WISC-III), and Wide Range Achievement Test Revised (WRAT-R) reading and arithmetic subtests, permitting the estimation of Verbal (VIQ), Performance, and Full Scale IQ. Handedness was assessed using the Edinburgh Handedness Questionnaire. Measures of cuπent psychopathology were obtained using the Young Mania Rating Scale (YMRS) and Global Assessment of Functioning (GAF) (APA, 1994). Antipsychotic doses (converted to chloφromazine equivalents) as well as number and type (antipsychotic, antidepressant, stimulant, anticonvulsant, lithium) of psychoactive medications at the time of scan were utilized as clinical variables. Structural imaging was performed at the McLean Hospital Brain Imaging Center on a 1.5 Tesla Scanner (Signa; GE Medical Systems, Milwaukee, WI). Acquisitions included a conventional Ti -weighted sagittal scout series (20 slices), a coronal T2 -weighted sequence to rule out gross pathology, a proton density/T₂-weighted interleaved double-echo axial series (116 slices, slice thickness = 3 mm, field of view (FOV) = 24 cm², TR = 3 sec, TE = 30/80 msec, acquisition matrix = 256 x 192, number of excitations = 0.5), and a 3- dimensional inversion recovery-prepped spoiled gradient recalled echo coronal series (124 slices, prep = 300 msec, TE = 1 minute, flip angle = 25°, FOV = 24 cm², slice thickness = 1.5 mm, acquisition matrix = 256 x 192, number of excitations = 2) which was used for structural analysis. All scans were reviewed by a clinical neuroradiologist to rule out gross pathology. Structural scans were transfeπed to the NMR Center for Moφhometric

Analysis (CMA)-Charlestown MGH and coded and catalogued for blind analysis. Imaging analysis was done on Sun Microsystems, Inc. (Mountainview, CA) workstations. Images were 'positionally normalized' to overcome variations in head position by utilizing a standard 3-dimensional brain coordinate system on each scan which used the midpoints of the decussations of the anterior and posterior commissure lines and the midsagittal plane at the level of the posterior commissure as points of reference for rotation and translation. This is a 'self-referential' system based directly on the individual brain, not waφed to a template atlas. The datasets were then segmented into gray, white, and CSF tissue classes using a semi-automated intensity contour algorithm for external border definition and signal intensity histogram distributions for delineation of gray-white borders. This technique allows for border definition as the midpoint between the peaks of the bimodal distribution for any given structure and its suπounding tissue. Segmentation of the structures or regions of interest (ROIs) were performed following the anatomic definitions of Filipek and colleagues for the total cerebrum, and of Seidman and colleagues for the thalamus (Filipek et al.(1994) Cereb Cortex 4, 344-360). The hippocampus was segmented following a procedure described by Seidman and colleagues (Seidman et al. (1999) Biol Psychiatry 46, 941-954; Seidman, et al. (2002) Arch Gen Psychiatry 59, 839-849). The segmentation of the amygdala has been performed manually in its entirety that comprised approximately 11 subsequent coronal sections. Using the cross-referencing capability of the program cardviews (Caviness et al, 1996), outlines deliminating the amygdala in axial and sagittal views were originally traced. This preliminary procedural step allows a reliable separation from the amygdala from suπounding gray structures such as the ventral putamen, the medial temporal cortex, and the hippocampus. The co-existence of amygdala and hippocampus in several coronal sections could make difficult the precise identification of the ventral amygdalar border (Kates et al 1997). The quality of the T-l weighted images used in the present study as well as the outlines traced on the cross-referenced axial and sagittal slice allow the visualization and segmentation of the amygdala reliably, therefore no conventions were needed for the definition of either the anterior amygdalar boundary (Altshuler et al 2000, DelBello et al (2004) Bipolar Disord.6(l):43-52.) or the amygdala-hippocampal junction (Altshuler et al 2000). The anterior most portion of the amygdala was segmented as it appears beneath the medial temporal cortex. At this region, the medial temporal cortex and the amygdala could give the impression of the thickening of the medial temporal cortex as has been reported previously( Altshuler et al 2000, DelBello et al 2004). The definition of this borders has been aided by the tracing of cross-referenced outlines in axial and sagittal planes. Superiorly the coroidal fissure has been used as the border of the amygdala along with the grey white matter contrast between the amygdala and the suπounding white matter. Similarly, the gray white matter contrast between the amygdala and its suπounding temporal white matter as well as the gray-CSF contrast between the amygdala and the temporal horn of the later ventricle has been considered as the lateral border of the amygdala. The parahippocampal cortex anteriorly and hippocampus posteriorly have been assign and the medial borders of the amygdala. Finally, the inferior border consisted of the gray- white matter contrast between the amygdala and its suπounding temporal white matter anteriorly and by the alveus and the temporal horn of the lateral ventricle posteriorly. These anatomical borders were visualized clearly by the contrast offered in the T-l weighted image used in the present moφhometric analyses. The volume for each structure was derived by multiplying the number of voxels assigned to each structure on each slice by the slice thickness followed by summing across all slices in which the structure appeared. There were no significant differences between youths with BPD and healthy controls in amygdala volumes, although the adjusted means for the R and L amygdalar volumes in the BPD youth were 5.6% smaller relative to those volumes in controls. Overall, boys had significantly larger R (t = -2.1, p < 0.05) and L (t = -2.8, p < 0.01) amygdala volumes. The R and L hippocampus were smaller in youths with BPD and this effect was more pronounced in girls than in boys (Sex*Dx, R: t = 2.2, p < 0.05; L: t = 2.3, p < 0.06). These volumes were 10% and 7.7% smaller than the R and L hippocampal volumes in the healthy controls. There were no differences between the controls and BPD youth in thalamic size, although the mean adjusted volumes of the R and L thalamus were 4.8% and 4.9% smaller than those volumes in controls. Boys had significantly larger thalamic volumes than girls (R: t = -4.1, p < 0.01; L: t = -4.8, p < 0.01). The BPD children had smaller TCV (5.4%) than healthy controls after adjustment for differences in age and head circumference. Both the R and L hemispheres were smaller, and this effect was seen in interaction with sex (R: t = 2.1, p = 0.04; L: t = 2.4, p = 0.02). No significant coπelations were seen between any clinical variables and either the hippocampus or TCV for the entire bipolar group. However, there was a trend for left hippocampal volume to vary inversely with YMRS score (r = -0.3, p < 0.1), which is suggestive of a relationship between having more symptoms of mania and a smaller left hippocampus. Interestingly, when the BPD children were separated by sex and the coπelation was rerun between the YMRS and left hippocampal volume, there was a significant relationship found in girls only (r = -0.5, p = 0.05). When the BPD group was assessed using the exploratory MANOVA analyses, we found no significant effects of age group, mood state, medication type, or presence of ADHD or psychosis on volume in the hippocampus or cerebrum. In conclusion, the children in this sample suffering from BPD had significantly smaller hippocampal and total cerebral volumes. OTHER EMBODIMENTS It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

WHAT IS CLAIMED IS:

1. A method of evaluating a subject, the method comprising: evaluating a region of the subject's brain using a moφhometric parameter, wherein the region includes the subject's amygdala, and the moφhometric parameter describes at least one structural feature of the amygdala; and determining whether the amygdala of the subject is altered relative to a reference amygdala.

2. The method of claim 1, wherein the reference amygdala represents an amygdala of a non-substance addicted subject and the alteration is a statistically significant reduction in the moφhometric parameter relative to the reference, and further comprising recommending a treatment or behavior to the subject, to avoid, prevent, or delay substance addiction.

3. The method of claim 1, wherein the reference amygdala represents an amygdala of a substance addicted subject and the alteration in the moφhometric parameter is not statistically significant relative to the reference, further comprising recommending a treatment or behavior to the subject, to avoid, prevent, or delay substance addiction.

4. The method of claim 2, wherein the recommended treatment or behavior facilitates avoiding, preventing, or delaying cocaine addiction.

5. The method of claim 1, wherein the reference amygdala is (i) an amygdala of a normal or control subject or (ii) a virtual structure whose structural features are based on structures of amygdalae of a plurality of subjects and a pre-selected probability value, the value representing a probability that a brain structure of one of the members of the plurality is within a constraint of the virtual brain structure.

6. The method of claim 1, wherein determining comprises comparing surface contours or volume of the amygdala of the subject to surface contours or volume of the reference amygdala.

7. The method of claim 1, wherein determining comprises evaluating undercut or overcut at one or more positions on the surface contour of the amygdala of the subject relative to surface contours of the reference amygdala.

8. The method of claim 1, wherein evaluating a region of the brain of the subject comprises magnetic resonance imaging (MRI); voxel-based moφhometry; segmentation-based moφhometry; or evaluating one or more surface features of the brain structure.

9. The method of claim 1, wherein evaluating a region of the brain of the subject comprises locating surface features of a brain structure in the region and comparing one or more of the surface features to an isoform surface for the brain structure based on the brain structure in each member of a cohort of subjects.

10. The method of claim 1, wherein the subject is treated with a candidate therapeutic agent or a candidate therapy, prior to, during, or after imaging the subject to obtain the moφhometric parameter.

11. A method of evaluating a subject, the method comprising: evaluating a brain structure of the subject to obtain moφhometric information about the brain structure; comparing the moφhometric information about the brain structure to coπesponding reference information, wherein the reference information is a statistical measure of a respective brain structure in each individual of a cohort of individuals.

12. The method of claim 11, wherein the moφhometric information describes volume or surface topology of the brain structure.

13. The method of claim 11. wherein each individual of the cohort is characterized by a behavioral disorder, a neuropsychiatric disorder, or a neurodegenerative disorder, or an abnormal performance in a behavioral paradigm.

14. The method of claim 11, wherein each individual of the cohort is or was addicted to an addictive substance.

15. The method of claim 14, wherein the substance is cocaine.

16. The method of claim 11, wherein the reference information is obtained by a method comprising: imaging at least a region of the brain that includes the brain structure in each individual of the cohort; aligning images obtained from each individual; and defining an isoform surface for the brain structure that is based on a preselected probability value, the value representing a probability that the brain structure of one of the members of the plurality is within the constraint of the virtual brain structure.

17. A system comprising a processor, a memory, and a communications interface, wherein the communications interface is configured to receive imaging information from an imaging apparatus, the processor is configured to process the imaging information according to the method of claim 11, and store results of the processing in the memory.

18. An article of machine readable media, having encoded instructions capable of causing a processor to effect the method of claim 11.

19. A method of evaluating a candidate therapeutic agent or candidate therapy, the method comprising: administering the candidate agent to a subject or providing the candidate therapy to the subject; and evaluating a brain structure, wherein a moφhological change in the brain structure indicates the candidate agent or candidate therapy is a lead compound or therapy for altering a trait.

20. The method of claim 19, wherein the trait is a behavioral trait..

21. The method of claim 19, wherein the brain structure is an amygdala.

22. The method of claim 19, wherein the behavioral trait is substance addiction.

23. The method of claim 22, wherein the substance addiction is cocaine addiction.

24. The method of claim 22, further comprising evaluating the test agent in a non-human animal model for addiction.

25. The method of claim 19, wherein the evaluating comprises voxel- based moφhometry or segmentation-based moφhometry.

26. The method of claim 19, wherein the evaluating comprises locating surface features of the brain structure and comparing one or more of the surface features to an isoform surface for the brain structure based on the structure in each member of a cohort of subjects.

27. The method of claim 19, wherein a cohort of subjects is used.

28. The method of claim 27, wherein the evaluating comprises determining over- or under-cuts in an iso-surface for a first cohort of subjects relative to a coπesponding iso-surface for a second cohort of subjects.

29. The method of claim 27, wherein the iso-surface is defined by a probability function.

30. The method of claim 27, wherein the cohort of subjects consists of subjects characterized by at least one common behavioral trait, and results of evaluating the cohort are compared to coπesponding results from a cohort of control subjects that do not have the common behavioral trait.

31. The method of claim 30, wherein the common behavioral trait is substance addition.

32. The method of claim 30, wherein the common behavioral trait is particularly cocaine addiction.

33. A method of evaluating a subject for a behavioral response associated with substance abuse, the method comprising: monitoring a subject during a reward-aversion paradigm, detecting a deviation in the subject's response relative to a normal or control subject, wherein the deviation is characteristic of a substance-addicted subject; and reporting the result of the detecting.

34. The method of claim 33, wherein the deviation is characteristic of cocaine addiction.

35. The method of claim 33, wherein the reward-aversion paradigm is a negative-outcome based monetary-reward paradigm and the deviation is characterized by absence of a biphasic response.

36. A method of providing a population-based statistic for a brain structure, the method comprising: evaluating images of a brain structure for each of a plurality of subjects; aligning the images; and determining positional information defining a virtual brain structure whose structural features are based on a pre-selected probability value, the value representing the probability that the brain structure of one of the members of the plurality is within the constraint of the virtual brain structure.

37. The method of claim 36, wherein the positional information represents an isoform surface.

38. The method of claim 36, wherein each subject of the plurality has a common behavioral trait that differs from normal, a common experience, a common learned ability, the same gender, or age.

39. The method of claim 36, wherein each subject of the plurality is addicted to a substance.

40. The method of claim 36, wherein each subject of the plurality has an abnormal characteristic in a behavioral paradigm.

41. The method of claim 36, wherein the virtual brain structure is smaller than normal.

42. The method of claim 36, wherein the brain structure is an amygdala.

43. The method of claim 36, wherein the aligning comprises locating one or more midpoints of decussations of the anterior and posterior commissures and the midsaggital plane.

44. The method of claim 36, wherein determining positional information comprises evaluating gray/white matter segmentation.

45. The method of claim 36, further comprising providing an estimate of risk for a behavioral trait, wherein the plurality of subjects each have a common behavioral trait.

46. A data-structure comprising moφhometric information about an amygdala, wherein the moφhometric information is a function of a statistic of a cohort of individuals with a common behavioral trait.

47. The data-structure of claim 46, wherein the moφhometric information comprises information about a volume of the amygdala or information about the surface topology of the amygdala.

48. The data-structure of claim 46, wherein the information about surface topology comprises information about a degree of undercutting or overcutting relative to a reference individual or a reference cohort, information describing at least a part of the amygdala, or information about the surface contours.

49. A database comprising a plurality of records, wherein each record of the plurality comprises (i) information that identifies a subject or personal or clinical information about the subject and (ii) moφhometric information that comprises at least one moφhometric parameter or a parameter that is a function of at least one moφhometric parameter.

50. The database of claim 49, wherein the moφhometric parameter comprises (i) a value representing a difference between an amygdala surface position in the subject relative to an isoform surface for a cohort, (ii) a value a difference between amygdala volume of the subject and a reference, or (iii) a scalar value that is a function of at least two moφhometric parameters.

51. The database of claim 49, wherein at least one record describes a subject who is addicted to an addictive substance.