DE102012202500A1 - Digital X-ray detector for receiving X-ray image irradiated by object, has sensor surface that is provided with uniform matrix structure having constant pixel pitch, and opening is formed between adjacent detector or sensor modules - Google Patents

Abstract

The X-ray detector has four-sided attachable rectangular detector modules (24) or sensor modules which are provided with x-ray converter that is divided into matrix of several pixels (12). The detector modules or sensor modules are arranged adjacent to one another on a common carrier. The sensor surface is provided with uniform matrix structure having constant pixel pitch (53). An opening (54) in sensor surface, is formed between adjacent detector modules or sensor modules. An electrical contact trans silicone vias is arranged in the sensor modules. An independent claim is included for method for correcting X-ray image.

Description

The invention relates to a digital X-ray detector according to claim 1 and to a method for correcting an X-ray image generated by the X-ray detector according to claim 10.

For diagnostic examination and interventional procedures, eg in cardiology, radiology and surgery, X-ray systems are used for imaging. X-ray systems 16 , as in 2 shown have an x-ray tube 18 and an X-ray detector 17 on, for example, together on a C-arm 19 arranged, a high voltage generator for generating the tube voltage, an imaging system 21 (often including at least one monitor 22 ), a system controller 20 and a patient table 23 , As X-ray detectors, X-ray flat detectors are generally used in many fields of medical X-ray diagnostics and intervention, for example in radiography, interventional radiology, cardiac angiography, but also imaging and imaging therapy in the context of control and radiation planning or mammography. Today's X-ray flat panel detectors are generally integrating detectors and are based primarily on scintillators whose light is converted into electric charges in arrays of photodiodes. These are then usually read out line by line via active controls. 1 shows the basic structure of a currently used indirect-converting X-ray flat detector, comprising a scintillator 10 , an active readout matrix 11 made of amorphous silicon with a multitude of pixels 12 (with photodiode 13 and switching element 14 ) and drive and readout electronics 15 (see eg M. Spahn, "Flat detectors and their clinical applications", Eur Radiol. (2005), 15: 1934-1947 ).

Recent research has focused on "counting" x-ray detectors, which individually count the incident x-ray quanta, rather than integrating them altogether, so that electronic noise can be almost completely suppressed. Counting X-ray detectors must, among other things, be able to deal with very high quantum fluxes of up to 10 ⁸ / s / mm ² and more, depending on the application. An additional quantification of the energy of each individual X-ray quantum results in further possible applications, for example those of material-specific imaging. Suitable materials for counting detectors are based on so-called direct converters using semiconductors such as CdTe, Cd (Zn, Te), CZT, HgI, PbO, etc. With these materials, an absorbed X-ray quantum directly generates electron-hole pairs, which are measured via an applied voltage and lead via a suitable read-out electronics to a counting event. The materials mentioned have advantages and disadvantages with regard to the required properties (such as absorption, drift velocity, etc.). The direct converter of the X-ray detector is generally structured in the form of pixels and is connected via a connection technology to an application specific integrated circuit (ASIC), for example, from CMOS technology, which also has a suitable pixelated structure.

Many of the direct converters that promise high signals and count rates, such as CdTe or CZT, can be produced with reasonable effort only in small areas, eg 2 × 2 cm ² or 3 × 3 cm ² . Even ASICs with a complex pixel structure, as required for counting detectors, can be produced with reasonable yield only in small areas. With greater effort slightly larger areas can be achieved, for example 2 × 8 cm ² or 3 × 6 cm ² , so that, for example, four 2 × 2 cm ² or two 3 × 3 cm ² large detectors can be applied to the corresponding ASICs. In any event, such detector modules are still small compared to the overall size of an average flat panel detector needed for angiographic applications (eg, 20 x 20 cm ² or 30 x 40 cm ² ). The detector modules can manage with or without a so-called guard ring around the detector edge. A guard ring defines the properties at the detector edges and, if appropriate, improves the properties of pixels arranged at the edge. In order to obtain a sufficiently large X-ray detector, therefore, a plurality of detector modules have to be lined up side by side or in a matrix-like manner (in the case of rectangular / square detector modules on 4 sides). In the following, a distinction is made between detector modules (only the detector material and the pixelated electrical contacts, possibly with guard ring) and sensor modules (detector module plus ASIC including connection technology, eg bump bonding between detector module and ASIC and TSV for connecting the ASIC to the peripheral electronics).

For mechanical and thermal reasons (accuracy with which the modules can be made, expansion, etc.), inadvertently a gap is created between modules stacked together. This suffers from the quality of the resulting X-ray images, as there are discontinuities. Another problem arises because electrical contacts not laterally but vertically with the help of so-called TSVs (trans silicon vias) through the silicon of the ASIC (application specific integrated circuits) chips must be led out, whereby the electrical connection (power supply of the chips, control and Data lines) with the underlying electronics is ensured. However, such TSVs again require vertical Openings, which are depending on the silicon thickness, for example, 100 microns to 200 microns.

It is an object of the present invention to provide a digital X-ray detector which enables an improved quality of X-ray images in counting X-ray detectors made from said materials. Furthermore, it is an object of the invention to provide a method for producing an X-ray image.

The object is achieved by a digital X-ray detector according to claim 1 and by a method according to claim 10. Advantageous embodiments of the invention are the subject of the dependent claims.

The digital X-ray detector according to the invention for acquiring an X-ray image of an X-ray irradiated object comprises a multiplicity of rectangular detector modules or sensor modules which can be arranged on four sides, wherein each detector module or sensor module has an X-ray converter and is subdivided into a matrix having a multiplicity of pixels the detector modules or sensor modules are arranged adjacent to one another on a common carrier, wherein the sensor surface formed by the entirety of the detector modules or sensor modules has a uniform matrix structure with a constant pixel pitch and a gap is arranged in the sensor area between adjacent detector modules or sensor modules. By virtue of the X-ray detector according to the invention, in particular a counting X-ray detector with X-ray converters designed as a direct converter, it is possible to produce high-quality X-ray images of an object, the advantages of new direct converter materials such as CZT and CdTe also being usable for large-area applications. Discontinuities in the X-ray image due to unwanted irregular gaps between detector modules may be caused by the regular pixel structure with a deliberately provided gap which fits into the regular matrix structure, in particular e.g. the size of a pixel, easily avoided. In addition, openings for possibly present TSVs can be arranged in the area of the gap, so that the successful use of a counting X-ray detector by the invention becomes possible in the first place. A pixel pitch is understood to be the distance between the centers of adjacent pixels.

According to one embodiment of the invention, the width of the gap between adjacent detector modules or sensor modules is designed in such a way that, while maintaining the pixel pitch, exactly one pixel is missing. On the one hand, such a gap is sufficient to provide openings for TSVs and, on the other hand, is not so large that small structures of the examination subject such as tumors or blood vessels could be overlooked. In each case, exactly one row or one column of pixels is missing between the detector modules or sensor modules, which are identified, for example, by suitable correction algorithms, e.g. in conjunction with interpolation, can be supplemented in the X-ray image.

It can also be provided that the width of the gap between adjacent detector modules or sensor modules is designed in such a way that at least two pixels are missing while maintaining the pixel pitch. This then corresponds to at least two rows or columns of missing pixels that need to be supplemented.

In an advantageous manner, a sensor module in each case has at least one detector module and one ASIC. In this case, the detector module has in particular a (mostly planar) upper electrode, the X-ray converter and pixelized lower electrodes (pixel electrodes), possibly additionally a so-called guard ring. The sensor module has an ASIC in addition to the detector module (s), in which case additionally connection technology such as e.g. Bump bonds can be present. The ASIC may e.g. be made using CMOS technology.

According to a further embodiment of the invention trans silicon vias (TSV) are arranged on the sensor modules for electrical contacting. Through the TSVs, electrical contacts of the ASIC are led out vertically through its silicon material, this being realized by means of openings in the silicon. The TSVs are advantageously arranged in the region of the gap.

According to a further embodiment of the invention, the pixels arranged on the edge of the detector module or sensor module at least on one edge side of the detector module or sensor module have a reduced pixel area of at least 15%, in particular also at least 30% or at least 50%, compared to the remaining, equally sized pixels and a distance between the normal sized pixels of a first detector module or sensor module arranged directly behind the reduced pixels and the normal sized pixels of a second detector module or sensor module arranged directly behind the reduced pixels such that while maintaining the pixel pitch at least two, in particular exactly two, fit in normal size pixels. In this embodiment, the edge pixels of each detector module are thus significantly smaller than the other pixels, but the pixel pitch, which depends on the "normal-sized" pixels, is maintained. By reducing the edge pixels, the gap between the Formed detector modules. The advantage is that here measured values do not have to be generated completely virtually but can be extrapolated from the measured values of the reduced pixels and thus are much more precise.

According to a further embodiment of the invention, the X-ray detector is formed by a flat-panel detector. Such a flat panel detector may e.g. in X-ray systems for interventional procedures e.g. used in cardiology, radiology and surgery. Frequently, such flat panel detectors are arranged together with an X-ray tube on a movable C-arm. Besides being used as a flat-panel detector, the X-ray detector according to the invention can also be used as a curved line detector, e.g. used in computed tomography.

Advantageously, for a simple and regular matrix structure, the detector modules are formed square. They may also be rectangular or have another shape that ensures a regular matrix structure (e.g., hexagonal with hexagonal pixels).

In addition, a method is also claimed for the correction of a tube X-ray image which was recorded with an X-ray detector according to the invention, wherein an X-ray image is generated by adding or correcting measured values for the region of the gap, which measured values are obtained by interpolation of the measured values of adjacent pixels or by extrapolation of the measured values be determined smaller pixels. For the X-ray detector, in which exactly one pixel is missing between the detector modules, a measured value is thus determined by interpolation between the measured values of the two corresponding edge pixels of the adjacent detector modules and supplemented as a correction of the tube X-ray image. This is done for all missing rows and columns. For the X-ray detector with the reduced pixels at the edges of the detector modules, the measured values obtained from the reduced pixels are corrected, the measured values being calculated by extrapolation. Advantageously, the measured values of the reduced pixels are made such, e.g. as a percentage of its size, extrapolated to fit a reading of a normal-sized pixel. In addition to the size, other properties of the pixels can also be included in order to carry out the extrapolation. After the corrections, a corrected X-ray image is obtained.

The invention and further advantageous embodiments according to features of the subclaims are explained in more detail below with reference to schematically illustrated embodiments in the drawing, without thereby limiting the invention to these embodiments. Show it:

1 a view of a known X-ray detector with a scintillator,

2 a view of a known X-ray system for use in interventional procedures,

3 a top view of a section of an inventive X-ray detector with multiple detector modules,

4 a plan view of an array of four detector modules according to 3 .

5 a plan view of a further arrangement of four detector modules according to 3 with guard rings,

6 a side view of a sensor module of an embodiment of the inventive X-ray detector,

7 a side view of two adjacently arranged sensor modules of an embodiment of the inventive X-ray detector,

8th a top view of four adjacently arranged sensor modules of an embodiment of the inventive X-ray detector,

9 a top view of an arrangement of four detector modules of a further embodiment of the inventive X-ray detector,

10 a top view of an arrangement of four detector modules of a further embodiment of the inventive X-ray detector,

11 a side view of two adjacently arranged sensor modules of a further embodiment of the inventive X-ray detector,

12 a side view of a sensor module with two detector modules of an embodiment of the inventive X-ray detector,

13 a plan view of the sensor module according to 12 .

14 a representation of the central functional elements of a counting pixel of an inventive X-ray detector,

15 a representation of the central functional elements of a counting and energy-discriminating pixel of an inventive X-ray detector and

16 a representation of an image processing chain in a recording of an X-ray image with an X-ray detector according to the invention.

In the 3 is a plan view of some detector modules 24 a counting digital X-ray detector according to the invention shown. Each detector module is in a plurality of equal-sized, square pixels 12 divided, which are arranged like a matrix and a regular pixel pitch 53 exhibit. The detector modules are arranged such that the regular matrix structure has no interruption, with a gap 54 is arranged between the detector modules. As in 4 shown by the pixels indicated by dashed lines, the gap is 54 designed to miss a pixel or a row or column of pixels. In the 5 is an array of four detector modules as in 4 shown, wherein the detector modules additionally each have a guard ring 25 exhibit. Guard rings can improve the properties of edge pixels. The detector modules shown have a 4-sided alignability, so that the x-ray detector can have any number of such detector modules next to each other. Correct "edge modules" may differ from the detector modules shown.

In the 6 is the structure of a sensor module 40 an X-ray detector according to the invention with a detector module 24 and a read-out ASIC 32 shown. The detector module has an upper electrode 29 , which can be designed flat, a direct converter 26 and pixel electrodes 27 on. If necessary, also a guard ring 25 at the detector module 24 be arranged. The direct converter 26 iA is not subdivided into a pixel structure, this is only due to the arrangement of the pixel electrodes 27 on the bottom of the direct converter 26 educated. Top here is the X-ray facing side and underside denote the side facing away from this. The direct converter can be formed, for example, from the materials CZT, CdTe, Cd (Zn, Te), HgI, PbO or similar semiconductor materials. These materials directly generate electron-hole pairs from an absorbed X-ray quantum. Next to the detector module 24 has the sensor module 40 a readout ASIC 32 and a number of trans silicon vias, TSVs, corresponding to the application 31 , on. The Elite ASIC 32 is via bump bonds 28 and electrical contacts 35 with the pixel electrode 27 the detector module conductively connected. Since the integrated circuit requires the entire area of a pixel, the TSVs can not be located within a sensor area of a pixel.

In the 7 are as cutting two sensor modules 40 a counting X-ray detector according to the invention shown, wherein between the two detector modules 24 the sensor modules 40 a gap 54 is arranged. The gap 54 corresponds to a defect of a pixel and the pixel pitch 53 remains over the individual sensor modules 40 respected. The sensor modules 40 are together on a substrate 33 arranged, for example of amorphous silicon. The substrate has one or more mechanical connections 38 to an underlying peripheral electronics board 34 on. In the area of the gap 54 the detector modules or their pixels 12 are the TSVs 31 the sensor modules 40 arranged, they lead through one or more openings 39 through the substrate and with electrical contacts 36 to the electronics board, being in the opening 39 electrical connections 37 are arranged. By such an arrangement absorbed X-ray quanta are converted into electrical charge in the direct converter, separated by the upper electrode and the pixel electrodes and pixelated transported via the bump bonds to the ASIC, the charge is collected in the ASIC and read and then further processed in the electronics board. Between the top electrodes 29 the sensor modules 40 are also HV connections 30 arranged.

In the 8th is a plan view of four juxtaposed sensor modules shown, in particular, the arrangement of the ASICs 32 and the TSVs on it 31 is visible. detector modules 24 and ASICs 32 are slightly shifted against each other and do not have to have exactly the same footprint. That's how the ASICs last 32 each into the spaces between two detector modules 24 into it. That's exactly where the TSVs are laid. By this arrangement, the regularity of the pixel structure can be obtained, and at the same time, the necessary electrical connections can be positioned one constraints.

The missing pixel columns or pixel lines created by the gap between the detector modules can be easily supplemented after taking up the raw data of the blank images of the individual sensor modules. The overall process of image processing is below in 15 described. However, after a raw picture 55 from the large number of blank images of the individual detector modules 24 is obtained, the missing pixel columns and rows can be determined or calculated by interpolation of the measured values of the pixels adjacent to the defects. Correction algorithms of similar known defect corrections (eg from the Script US 6763084 B2 , if here also dynamic and not primarily stationary occurring defects are treated) are used.

In the 9 and 10 is as a further embodiment of the invention is a plan view of four detector modules 24 , which each have smaller pixels at their edges 12.1 have shown. The reduced pixels 12.1 are at least 20% smaller in their sensor surface than the other normal sized pixels 12 In particular, the respective corner pixel is formed even smaller because it is adjacent to two other detector modules. For four-sided alignability, the pixels at all four edges of the respective detector modules are reduced in size. The detector modules 24 are arranged in such a way and the gap is formed such that the pixel pitch is maintained and the reduced pixels 12.1 are arranged at the location of a normal sized pixel. This is shown by means of the dashed lines indicated pixels. 10 shows a similar arrangement as 9 with the difference that the detector modules each have a guard ring 25 exhibit.

11 shows the arrangement of 9 and 10 as a side view and taking into account the complete sensor modules 40 and their arrangement on the substrate 33 , In the area of the gap 54 In turn, the TSVs are arranged. The reduced pixels 12.1 are mainly due to a reduction in the area of the reduced pixel electrodes 27.1 educated. The reduction of the sensor area results in a reduced sensitivity of the reduced pixels 12.1 , However, this can be easily corrected in the gross overall picture.

The lower sensitivity or lower count rate over a full sensor area pixel is highly corrected according to the percentage of reduced sensor area, e.g. extrapolated. Calibration data can also be used for this: With uniform irradiation (except for statistical fluctuations), the same count rate can be expected in the reduced pixels as in normal-sized pixels. In fact, however, a different, lower count rate is counted because of the lower effective detector area of these edge pixels, but also because of other physical effects, e.g. k-Escape, Comptonstring, other manifestation of the electric field that collects the charge carriers, etc. The discrepancy between the expectation and the measured count rate in edge or central pixels (calibration) can be used to compensate for the "wrong" count rate of the reduced pixels , The i.a. lower counts of the reduced pixels also produce higher relative noise due to the reduced quantum statistics in succession. After the count rates have been adjusted by the correction, the noise in the reduced pixels is higher than in the normal sized pixels. However, this increased noise is also deterministically predictable by the calibration process and can be compensated by a suitable noise reduction measure.

In the 12 and 13 an embodiment of a counting X-ray detector according to the invention is shown, in which a sensor module 40 two detector modules 24 side by side, which is on an ASIC 32 are arranged. The TSVs 31 are in turn located on the edges of the ASIC, leaving them in the gap area 54 can be arranged. Other combinations of n × m detector modules per sensor module are of course also possible, for example 3 × 1, 2 × 2, 3 × 2, 4 × 1, etc. detector modules per sensor module. In any case, in the arrangement of the detector modules on the sensor module, the pixel pitch is maintained, so that, for example, with detector modules with reduced pixels at the edges, there is a gap between the detector modules. The measured values can then be corrected again accordingly.

14 shows functional elements of an integrated circuit of a known counting pixel of a counting X-ray detector, which preferably also has an inventive X-ray detector. The charge input 41 is collected in the pixel and is using a charge amplifier 42 and a feedback capacity 46 strengthened. In addition, the pulse shape in a shaper (filter) can be adapted at the output (not shown). An event is then counted by a counter 44 is incremented by one if the output signal is above an adjustable threshold. This is about a discriminator 43 detected. The threshold is over eg via a DAC 56 (digital to analog converter). 15 shows functional elements of an integrated circuit of a counting pixel, which is also able to distinguish between several energy ranges of the X-radiation. For this purpose, four (the number can be chosen as needed) are different discriminators 43 present, which have different thresholds, so that by means of various counters 44 Events of different energy can be counted. A readout logic 46 evaluates the corresponding results.

16 shows the general steps of a correction method, which of the readout of the individual blank images 48 the individual detector modules or sensor modules to a final result X-ray image 47 leads. First, the read-out blank images 48 independent of a module-specific raw data correction step 49 subjected. Subsequently, the corrected blank images are stitched using a stitching algorithm. 50 merged into a raw total image. Such stitching algorithms are known from a variety of applications, for example, in more CCD detectors for radiography / mammography or satellite imagery or constellations, which together produce a larger view. Such an algorithm is eg in the US Pat. No. 6,718,011 B2 shown. The gross picture 55 is then in a correction step 51 As already described in detail above corrected by missing missing pixels supplemented or reduced pixels are extrapolated. In the case of missing rows or columns of pixels between detector modules, therefore, a restoration of the image content on the nonexistent pixels is performed similar to a defect correction. In the case of reduced pixels, correction of reduced pixel count rates and increased noise of the reduced pixels is performed based on previously generated calibration data. Subsequently, a clinical image processing step can be added 52 be done so that in the end the result x-ray image 47 is obtained.

The invention may be summarized in the following manner: For improved image quality in counting flat-panel detectors with direct converters, a digital X-ray detector for taking an X-ray image of an X-ray irradiated object comprising a plurality of rectangular detector modules or sensor modules which can be arranged on four sides, wherein each detector module or sensor module has an X-ray converter and is subdivided into a matrix having a plurality of pixels, wherein the detector modules or sensor modules are arranged adjacent to one another on a common carrier, the sensor area formed by the entirety of the detector modules or sensor modules having a uniform matrix structure with a constant pixel pitch and in each case a gap in the sensor surface is arranged between adjacent detector modules or sensor modules.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• US 6763084 B2 [0038]
• US Pat. No. 6718011 B2 [0044]

Cited non-patent literature

• M. Spahn, "Flat detectors and their clinical applications", Eur Radiol. (2005), 15: 1934-1947 [0002]

Claims (11)

Digital X-ray detector for recording an X-ray image of an object irradiated by an X-radiation, comprising a multiplicity of rectangular detector modules which can be arranged on four sides ( 24 ) or sensor modules ( 40 ), each detector module ( 24 ) or sensor module ( 40 ) has an x-ray converter and into a matrix with a plurality of pixels ( 12 . 12 .1), wherein the detector modules ( 24 ) or sensor modules ( 40 ) are arranged adjacent to each other on a common carrier, which by the entirety of the detector modules ( 24 ) or sensor modules ( 40 ) formed a uniform matrix structure with a constant pixel pitch ( 53 ) and between adjacent detector modules ( 24 ) or sensor modules ( 40 ) each have a gap ( 54 ) is arranged in the sensor surface. An X-ray detector according to claim 1, wherein the width of the gap ( 54 ) between adjacent detector modules ( 24 ) or sensor modules ( 40 ) is designed such that, while maintaining the pixel pitch ( 53 ) exactly one pixel can be inserted. An X-ray detector according to claim 1, wherein the width of the gap between adjacent detector modules ( 24 ) or sensor modules ( 40 ) is designed such that, while maintaining the pixel pitch ( 53 ) at least two pixels can be inserted. X-ray detector according to one of the preceding claims, wherein for an electrical contact trans-silicon vias ( 31 ) on the sensor modules ( 40 ) are arranged. X-ray detector according to one of the preceding claims, wherein the edge on the detector module ( 24 ) or sensor module ( 40 ) arranged pixels at least on one edge side of the detector module ( 24 ) or sensor module ( 40 ) has a pixel area reduced by at least 15% compared to the remaining, equally sized pixels ( 12 ) and wherein a distance between the directly behind the reduced pixels ( 12.1 ) arranged normal sized pixels ( 12 ) of a first detector module ( 24 ) or sensor module ( 40 ) and directly behind the reduced pixels ( 12.1 ) arranged normal sized pixels ( 12 ) of a second detector module ( 24 ) or sensor module ( 40 ) is designed such that, while maintaining the pixel pitch ( 53 ) at least two, in particular exactly two, pixels are insertable. An X-ray detector according to one of the preceding claims, wherein the X-ray detector is provided by a counting X-ray detector as a direct converter ( 26 ) formed X-ray converters is formed. X-ray detector according to one of the preceding claims, which is formed by a flat-panel detector. X-ray detector according to one of the preceding claims, wherein the detector modules ( 24 ) are square. X-ray detector according to one of the preceding claims, wherein a sensor module ( 40 ) at least one detector module ( 24 ) and an ASIC ( 32 ) having. Method for correcting a pipe X-ray image, which was taken with an X-ray detector according to one of claims 1 to 9, whereby an X-ray image is generated by specifying for the region of the gap ( 54 ) Measured values can be supplemented or corrected which measured values are obtained by interpolation of the measured values of neighboring pixels ( 12 ) or by extrapolation of the measured values of reduced pixels ( 12.1 ) be determined. Method according to claim 10, wherein in the case of reduced pixels ( 12.1 ) the measured values of the reduced pixels ( 12.1 ) are extrapolated in proportion to their size so that they correspond to a measured value of a normal-sized pixel ( 12 ) correspond.

Images