Photon counting detector flicker noise correction

Abstract

The present invention relates to a device (20) for photon counting image detection (e.g., to correct observed counts). The apparatus comprises an ASIC (32) for connection to a detector array (30) comprising detector pixel circuitry (34) for converting energy of photons into electrical packets. The apparatus comprises a processor (50) for processing an output of the application specific integrated circuit. The ASIC includes pixel circuits each adapted to receive an electrical input pulse, where the pixel circuits are substantially identical in function and / or structure. The ASIC comprises a pulse generator (39) for generating an electrical test pulse. The image pixel circuit is connectable to a corresponding detector pixel to receive its electrical input pulse, and the dummy pixel circuit is connected to a pulse generation circuit (39) to receive its test pulse as input. A processor (50) analyzes the dummy pixel circuit output to determine a metric indicative of drift, variation, and / or deviation of the at least one parameter. The processor also corrects the image pixel circuit output and / or adjusts the setting of the ASIC by taking into account the difference metric.

Description

Technical Field

[0001] This invention relates to the field of image detectors, and more particularly to the field of photon-counting image detectors, for example, for diagnostic imaging. More specifically, this invention relates to apparatus, systems, methods, and computer program products for characterizing, reducing, and / or compensating for flicker noise in signals and / or (e.g., images derived from such signals) acquired by a photon-counting detector. Background Technology

[0002] In X-ray radiation detectors, sensor elements generate charge when exposed to incident X-ray (and / or gamma) radiation. For example, indirect conversion sensors typically use a scintillator to convert X-ray radiation striking the sensor into short bursts of light, which are then converted into charge, for example, by means of a photodiode. In direct conversion sensors, X-ray photons interact directly with the sensor material to generate charge, which in principle allows for the individual detection of each photon interaction within practical constraints.

[0003] Photon counting (PhC) detectors are typically radiation detectors suitable for detecting and counting individual photons. For example, such a single-photon counting detector can emit a signal pulse for each detected photon, as opposed to a conventional (integrating) photodetector whose output is proportional to the total photon flux impacting the radiation-sensitive area (pixel). Counting detectors are commonly used in nuclear medicine (e.g., positron emission tomography (PET) and single-photon emission computed tomography (SPECT)) and can also offer various advantages for 2D X-rays and computed tomography (CT) (e.g., for spectral CT imaging and / or dynamic CT imaging).

[0004] Photon counting detectors typically count photons by comparing the incoming detector signal to an energy threshold, or, in the case of spectral imaging, to more than one energy threshold. The performance and achievable output quality of CT systems using PhC detectors are highly dependent on the temporal stability of the sensor and associated readout electronics.

[0005] In a typical detector design, the direct conversion material is coupled to an application-specific integrated circuit (ASIC) for converting radiation-induced charge into a per-pixel electrical signal. For example, a pixelated detector array (e.g., a CZT crystal) can be directly connected (and physically attached) to the ASIC via flip-chip bonding (but not limited to this; for example, an interpolator structure could also be used). The ASIC (e.g., an array of such ASICs) processes the raw signal (charge packet) originating from the radiation-sensitive element. This processing can be implemented, for example, using a pulse shaper and / or other analog processing, through one or more counting channels, and may include an analog-to-digital conversion (ADC) step.

[0006] Photon counting detectors are sensitive to flicker noise in the readout circuitry implemented on an ASIC. This flicker noise can correlate with the energy threshold level over time, causing observation counts (OCs) across different pixels to also be correlated. Since observation counts form the (raw) basis for photon flux estimation and are typically further processed into images of interest (e.g., CT reconstructed images expressed in Hounsfield units), this correlated noise can significantly contribute to unwanted image artifacts in reconstructed CT images (e.g., in tomographic slices). Additionally, flicker noise can affect the amplitude of the pulse at the output of the pixel's analog front-end (AFE), potentially leading to incorrect or inaccurate energy discrimination.

[0007] Even though recent MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) design optimizations have been designed to reduce flicker noise caused by the analog front end of the pixel and other shared ASIC structures, flicker noise remains a significant contributor to the instability of observation counts (OC) over time, in addition to other factors such as electric field variations in radiation-sensitive sensor materials (e.g., in direct conversion material crystals (e.g., CZT)).

[0008] The article “Photon Counting X-ray imaging with CdTepixel detecotrs based on XPAD2 circuit” by Romain Franchi et al. (Nuclear instruments & methods in physics research, Chapter A, Elsevier BV, North-Holland, Vol. 563, No. 1, July 1, 2006) discloses a pixel unit circuit that includes a test pulse injector connected to the input of an amplifier for inserting test pulses into the detector pixel amplifier.

[0009] Patent EP2517048B1 discloses a radiation detector assembly having signal processing circuit elements connected to corresponding detector pixels, wherein a test circuit is configured to inject test electrical pulses into the signal processing circuit. Summary of the Invention

[0010] The purpose of embodiments of the present invention is to provide an apparatus and method for reducing flicker noise in signals acquired by a photon counting detector.

[0011] This invention proposes monitoring the evolution of energy threshold levels and pulse amplitudes (e.g., due to flicker noise) over time in pixels on the same ASIC of a photon counting detector. Electrical test pulses can be directly injected into the analog front end of the (dummy) pixel circuitry on the ASIC, for example, replacing the original sensor pulses caused by the conversion of radiation (e.g., through photon interactions in a CZT crystal) into charge. Due to the correlated nature of flicker noise, threshold and amplitude variations in other pixel circuitry of the same ASIC (i.e., image pixels receiving actual sensor input) can be estimated using electrical test pulses with known characteristics (e.g., known amplitude).

[0012] The advantages of embodiments of the present invention are: good stability of the output of the photon counting detector (e.g., observation count (OC)) and / or reduced correlation between errors in the outputs related to different detector pixels (read from the same ASIC) due to flicker noise. The counting energy threshold level can be tuned more accurately and / or more robustly, thereby achieving better photon energy discrimination and reducing image artifacts due to spatial (or spatiotemporal) correlated noise.

[0013] This invention can be applied to various detector designs (e.g., regardless of specific radiation-sensitive materials and / or specific analog front-end circuit designs) and can also be used for photon counting detectors based on gamma radiation. It should be understood that, even though this specification primarily refers to direct-conversion PhC detectors, the same approach can be applied to photon counting detectors based on indirect detection (e.g., using intermediate scintillators and, for example, photodiode structures).

[0014] The advantages of embodiments of the present invention are that test signals with known characteristics (e.g., known amplitude) can be conveniently used for calibration and / or testing purposes, for example, without the need for sophisticated test and / or calibration setups, extensive instruments, and / or complex procedures.

[0015] This invention is implemented by the apparatus, system, method and computer program product described in the claims.

[0016] ASICs can be adapted to avoid signal inputs from the detector array to the dummy pixel circuitry. This can be achieved by, for example, electrically isolating the pre-fabricated electrical connections of one or more dummy pixels to the detector array (so that the pixel array can be manufactured as a uniform set of pixels, e.g., except for providing passivation on the connection pads of the dummy pixels); physically lacking detector pixels at the input nodes of one or more dummy pixels during mounting (e.g., the detector array does not extend over the area occupied by the dummy pixels, e.g., in the last row of the ASIC pixel matrix); designing the ASIC pixels such that the dummy pixels are identical to the image pixels (except for lacking input connections to the detector array), etc.

[0017] Analyze the output of at least one dummy pixel that responds to multiple electrical test pulses to determine a measure of drift, variation, and / or deviation of at least one parameter of the ASIC and / or detector.

[0018] This metric is used to correct the output of other image pixel circuits (e.g., threshold detectors) and / or to adjust the settings of a dedicated circuit (e.g., adjust the setting of at least one threshold of the image pixel circuit (threshold detector)) by taking the determined metric into account to perform count correction (CoC) on the measured image pixel signal, for example.

[0019] In the following text, the term "image pixel" will also be used to refer to "image pixel circuit".

[0020] Specifically, this metric can involve the change in the correspondence between a threshold level and the amplitude of an input pulse that exceeds the threshold due to electronic noise (e.g., flicker noise). In particular, electronic noise (e.g., flicker noise) may cause the effectively applied threshold (e.g., by a comparator of a threshold detector) to differ from its nominal setting, and / or may shift the actual input received by such a comparator to a different value than the input that should be received without such noise (i.e., the actual input signal or the signal derived therefrom as expected). Such noise, or at least its components, is of primary concern to this invention as being related across pixels on the same ASIC (i.e., applied to image pixels and (one or more) dummy pixels), and can generally be characterized as a slow drift (e.g., a low-frequency change in the correspondence (drift of the effective threshold and / or the amplitude of the signal compared to the threshold)).

[0021] The independent and dependent claims describe the specific and preferred features of the invention. Features of the dependent claims can be combined with features of the independent claims and other features of dependent claims deemed appropriate, but not necessarily only as expressly stated in the claims. Attached Figure Description

[0022] Figure 1 An illustrative CT imaging system according to an embodiment of the present invention is shown.

[0023] Figures 2a and 2b illustrate a device according to an embodiment of the present invention.

[0024] Figure 3 An illustrative pixel matrix layout of an ASIC in a device according to an embodiment of the present invention is shown.

[0025] Figure 4 An amplitude scan relative to a fixed threshold level is shown to illustrate an embodiment of the invention.

[0026] Figure 5 It is shown as such Figure 4 The curves of the output (observation count) of dummy pixels as a function of the pulse amplitude in amplitude scanning are schematically shown to illustrate embodiments of the invention.

[0027] Figure 6 A threshold scan is illustrated, in which a sequence of substantially the same (i.e., constant amplitude) pulses is used to test different threshold levels to illustrate an embodiment of the invention.

[0028] Figure 7 It is shown as such Figure 6 The curve of the output (observation count) of the dummy pixel as a function of the threshold level in a threshold scan is schematically shown to illustrate an embodiment of the invention.

[0029] Figure 8 Illustrative experimental data obtained from a detector array according to an embodiment of the present invention are shown, which show: image pixel output over time during (blank) image acquisition (observation count; averaged over a set of pixels), a filtered (smoothed) version of the aforementioned image pixel output, and output acquired simultaneously from dummy pixels (averaged over a set of dummy pixels and similarly smoothed).

[0030] Figure 9 Illustrative experimental data, as a function of time (index), is shown from image pixels during different blank image acquisitions (shown in different gray tones). (i.e., the average count of observations over multiple pixels).

[0031] Figure 10 This shows the result after applying correction based on the output of dummy pixels acquired simultaneously with the image pixel acquisition. Figure 9 The illustrative example of observation counts is used to illustrate the effectiveness of the correction method according to an embodiment of the invention.

[0032] Figure 11 The illustration shows a method according to an embodiment of the present invention.

[0033] The accompanying drawings are illustrative and not restrictive. Elements in the drawings are not necessarily shown to scale. The invention is not necessarily limited to the specific embodiments shown in the drawings. Detailed Implementation

[0034] The present invention relates to application-specific integrated circuits (ASICs) suitable for image detection based on (e.g., X-ray and / or gamma) photon counting (e.g., in computed tomography (CT) scanners with (photon counting, e.g., spectral photon counting)).

[0035] Figure 1 An illustrative diagnostic imaging system according to an embodiment of a third aspect of the invention is shown, comprising a photon-counting detector 110 used in a computed tomography (CT) system. Alternatively, the imaging system may be a conventional projection imaging system, such as a direct radiography system or a mammography system. The imaging system is not necessarily limited to projection radiography; for example, embodiments of the invention may also relate to or be applied to nuclear medicine (e.g., SPECT imaging or PET imaging) or non-medical fields (e.g., non-destructive testing, safety screening via radiography), etc. For example, as discussed below, the radiation detector assembly and related methods according to the embodiments may also be implemented (or combined with) a PET or SPECT radiation detector assembly, a gamma camera head, or another single-photon (or photon-counting) image detector. Nevertheless, it should be understood that embodiments of the invention may be particularly advantageous when applied to CT imaging detectors, and especially for applications such as spectral CT (e.g., spectral photon-counting computed tomography (SPCCT)).

[0036] refer to Figure 1 A diagnostic imaging system 100 according to an embodiment of the present invention includes a radiation source 108 and a radiation image detector 110 (i.e., a photon counting image detector according to an embodiment of the present invention (see further description below)). The imaging system 100 may include a generally stationary gantry 102 and a rotating (rotatable) gantry 104. The rotating gantry 104 may be rotatably supported by the fixed gantry 102 and may (in use of the system) rotate about a longitudinal axis Z around an examination area 106. The radiation source 108 and the radiation image detector 110 may be held by the gantry in a position opposite each other across the examination area 106, such that the radiation source 108 can project X-ray radiation through the examination area 106 to collide with the radiation image detector.

[0037] The radiation source 108 (e.g., an X-ray tube) can be rotatably supported by a rotating frame 104 so as to rotate together with the rotating frame, and can be adapted to emit X-ray radiation through the examination area 106.

[0038] The radiation image detector 110 may include a radiation-sensitive detector array positioned across the inspection area 106 at an angular arc relative to the radiation source 108. The image detector 110 may include one or more rows of detector elements arranged relative to each other along the Z-axis. The image detector 110 may be adapted to detect radiation passing through the inspection area 106 and generate a signal indicative of that radiation (e.g., an indication of the intensity of radiation received at different locations on the array or of a single photon).

[0039] The system may include an object support 113 (e.g., a movable examination table) for supporting a target or object in the examination area.

[0040] The CT system may include a reconstructor 112 for reconstructing signals provided by image detector 110. The reconstructor may be adapted to generate reconstructed tomographic images (e.g., showing transverse slices). The system may also include an operator console 114 for controlling or monitoring system 100 and / or for providing a user interface to the operator.

[0041] An apparatus according to an embodiment of this first aspect of the invention includes an integrated circuit (ASIC) and one or more processors and / or one or more controllers for, for example, processing signals and / or one or more other outputs of the ASIC, determining image data based on the signals and / or one or more other outputs of the ASIC, determining an image commensurate with the pixel layout of the detector, and / or controlling the operation of the ASIC. For example, the processor may be adapted to generate an image representing the photon flux exposed to the detector array module and / or one or more spectral images (e.g., related to a specific energy subdivision of radiation impacting the detector). The image data is not necessarily limited to this direct representation of radiation impacting the detector; it may also involve, for example, tomographic reconstruction and / or other image data derived from the original detector signals.

[0042] Referring to Figures 2a and 2b, which schematically illustrate an illustrative device 20 (e.g., a radiation detector assembly) according to an embodiment of the invention, device 20 includes an application-specific integrated circuit (ASIC) 32. The ASIC 32 is operatively connected to and / or adapted to be operatively connected to a detector array module 30, which is used to convert photons into electrical detection pulses, for example, to provide, in a spatially resolved manner, one or more electrical output signals representing the interaction of radioactive materials in the radiation-sensitive material of the detector array module (i.e., enabling the acquisition of image information and / or the inference of image information based on one or more of the electrical output signals). For PhC, the detector array module may generate a single electrical output signal (e.g., a pulse) in response to the interaction of a single photon with the detector array module (i.e., to allow photon counting and / or (i.e., energy detection of each) photon).

[0043] Device 20 may also include the detector array module 30, for example, ASIC 32 may be operatively connected to the detector array module 30. For example (but not limited to), ASIC 32 may be a flip chip bonded to the detector array module 30, for example, as schematically illustrated in FIG2 by means of structural and / or conductive connection points (e.g., solder balls).

[0044] The detector array module 30 includes a plurality of detector pixels 34 (e.g., an array of detector pixels 34). Detector pixels may include suitable direct conversion materials, such as a CdTe-ZnTe alloy system (e.g., CZT) for converting photons into charge packets (causing detectable electrical pulses). The detector array module may be pixelated into multiple pixels, for example, forming a matrix with rows and columns. Each detector pixel 34 typically includes a cathode and an anode to define the volume of charge collected by the pixel thereon from the radiation-induced charge, and routes the collected charge to the ASIC board via an interconnect between the ASIC and the module.

[0045] The ASIC is connected to the radiation-sensitive detector pixel interface via a direct (or at least preferably short and low-resistance) electrical connection between the detector pixel 34 and the input portion of the ASIC (e.g., the input node of the corresponding ASIC image pixel 35) to process the raw analog signal (collected charge packets) representing the radiative interactions within the radiation-sensitive volume of the detector module (i.e., within the physical range of each detector pixel). For example, in a flip-chip arrangement, charge can be transferred substantially directly from the detector pixel 34 to the corresponding ASIC image pixel 35, for example, via bonding connector 38.

[0046] The ASIC receives raw electrical input from the radiation-sensitive element (e.g., for each detector pixel 34) and generates one or more outputs to characterize and / or quantize (e.g., again for each corresponding pixel) the input. The image pixel 35 of the ASIC can implement generally known functions (e.g., amplification, pulse shaping, pulse counting, integration, thresholding, filtering, digitization, and / or other such functions for conditioning the raw electrical signal (e.g., charge) into one or more usable signals) to generate one or more outputs based on the raw analog input. It should be understood that not all processing operations for obtaining an image from the raw input pulses from the detector array must be implemented by the ASIC; for example, the ASIC may primarily provide the lowest level of processing functions (e.g., amplification (if needed / implemented), pulse shaping, thresholding detection (e.g., analog-to-digital conversion of pulse amplitude), and / or counting). Higher-level functions may be implemented on an ASIC, partially on an ASIC, or entirely delegated to one or more additional processing systems, referred to for simplicity as processors (and / or controllers) 50 and / or (e.g., equivalently referred to as) the image chain (IC) processes discussed further below.

[0047] The ASIC 32 includes a plurality of image pixels 35 connected to or adapted to be connected to a plurality of corresponding detector pixels 34 of one or more detector array modules 30. Thus, the plurality of image pixels 35 of the integrated circuit can be operatively connected to determine the radiation received by a plurality of corresponding radiation-sensitive volumes (e.g., CZT pixels) in one or more radiation-sensitive detector array modules 30. The image pixels 35 can form a one-dimensional or two-dimensional pixel array.

[0048] Each detector pixel 34 can be connected in use to a corresponding signal processing circuit (i.e., the signal processing circuit implemented by the image pixel 35), which is adapted to digitize the electrical pulses (charge packets) generated in the radiation-sensitive material. The image pixel may include a pulse shaper and / or other analog processing circuitry 41 to shape (and / or amplify) the raw pulses from the detector pixel into cleaner and / or more representative (e.g., a more Gaussian-like shape with a predetermined width) pulses (e.g., the pulse shaper is adapted to normalize the processing shape and full width at half maximum (FWHM) as optimally as possible). The regulated pulses can typically have a height at least approximately proportional to the photon energy of the X-ray photons that have already generated the electrical pulses in the detector pixel.

[0049] For example, after the pulse is shaped by the pulse shaper 41, the pulse can be fed to the threshold detector 42 (e.g., including one or more comparators). Multiple comparators with different threshold levels can be used for detection / counting in multiple energy bins (e.g., for photon energy discrimination and / or spectral imaging). The threshold detector can also be configured as a simple analog-to-digital converter (ADC) (e.g., by generating a quantized (i.e., digital) output representing the pulse amplitude (e.g., a 1-bit output for simple single-threshold detection in the absence of energy discrimination, and / or allowing the identification of multiple bits in different energy bins using different digital multi-bit values)).

[0050] By configuring comparators for all image pixels in the same way (e.g., implementing the same threshold), homogeneous image responses can be achieved within tolerances. Alternatively, the threshold (comparator level) can be configured per pixel for calibrating individual pixel circuitry.

[0051] Device 20 also includes a processor 50 (as indicated by arrows 44, one or more) for processing the outputs of the integrated circuit. The processor may be integrated into the ASIC 32, external to the ASIC 32, or a combination of both. The processor 50 may also be adapted to control the functions of the ASIC, such as controlling the pulse sequence generated for dummy pixels (discussed below), adjusting the threshold trimming level of the image and / or dummy pixels, etc. (as indicated by arrow 43).

[0052] Furthermore, the integrated circuit includes at least one dummy pixel 36, which may have the same or similar structure as the aforementioned image pixels. One or more dummy pixels are not connected to the radiation-sensitive detector array module 30. Either the bonding connection 38 is absent, or if it is present, it is not used. Instead, the dummy pixel receives raw input from the pulse generator circuit 39 (e.g., controlling the pulse generator circuit to act as an AFE input for the dummy pixel).

[0053] Figure 2a illustrates a single pulse generator circuit typically shared by one or more dummy pixels and optionally with regular image pixels used for calibration purposes. Figure 2b illustrates a case where each dummy pixel (and optionally, each image pixel) has its own pulse generator circuit 39. Of course, any solution among these is also possible, i.e., a finite number of shared pulse generator circuits. Having independent pulse generator circuits allows the processor to configure the pulse generator circuits individually and / or in parallel.

[0054] The dummy pixel 36 may include an analog front-end (AFE) circuit adapted to receive input from the pulse generator circuit 39. The image pixel 35 may include an analog front-end circuit of similar or identical structure. Optionally, the front-end circuit of the image pixel 35 may be adapted to receive input from the pulse generator circuit 39 to enable the pulse generator circuit to calibrate individual image pixels.

[0055] In one embodiment, at least one dummy pixel is implemented outside the image pixel matrix. Obviously, the dummy pixel(s) may also be located within the area covered by the detector area module (and therefore by one or more detector pixels), but configured to lack electrical connection to the detector area module (and therefore, one or more detector pixels). This advantageously allows the dummy pixel(s) to be implemented with substantially the same design and components as the image pixels (e.g., only with smaller adapters to provide different functionality for the dummy pixel). Figure 3 An example of a pixel matrix layout is shown, in which dummy pixels occupy the first column of the matrix. Embodiments are not necessarily limited to this example; for example, the first row, the last row, the first column, the last column, and / or any combination thereof can be used for dummy pixels. Dummy pixels are not necessarily limited to a single first row and / or column and / or a single last row and / or column; for example, the first (and / or last) 2, 3, ... rows and / or columns can be used for dummy pixels. Even so, it should be understood that the number of rows and / or columns allocated to dummy pixels is preferably kept low, for example, less than 10% of the total number of pixels, preferably less than 5%, for example, about 1% or even less. It should also be noted that an entire row and / or column does not necessarily have to be filled with dummy pixels; for example, only one or a few pixels in a row and / or column may be sufficient to serve as dummy pixels.

[0056] Alternatively, the detector pixels connected to the dummy pixels can be made inactive (e.g., by shielding a portion of the detector array so that no photons can reach the radiation-sensitive material of the detector pixels connected to the dummy pixels). It should be understood that this method may not be ideal, but it is easy to implement.

[0057] By positioning one or more dummy pixels outside the image pixel matrix but within the same ASIC, any structural differences between the dummy pixels and the image pixels can be minimized without interfering with normal detector operation, such as allowing the center field of the image pixels to acquire data for each pixel location in the image pixel grid. By limiting structural differences, the flicker noise effect in the dummy pixels will be substantially similar to or identical to the flicker noise effect in the image pixels, such that the characterization of flicker noise determined via the dummy pixels can be assumed to represent the flicker noise experienced by the actual image pixels on the same ASIC.

[0058] The integrated circuit includes a pulse generator circuit 39, which is adapted to generate a plurality of predetermined charge pulses, for example, each pulse corresponding to a fixed amount and / or a configurable amount of charge (e.g., typically a known charge, and preferably a precisely controllable charge). The pulse generator circuit 39 may be included in each dummy pixel, or each dummy pixel may be connected to one or more common pulse generator circuits external to the dummy pixel.

[0059] Pulse generator circuit 39 is configured to inject electrical pulses into the signal processing circuitry of dummy pixel 36. Similar or identical pulse generator circuitry may also be provided to inject such electrical pulses into regular image pixels 35, for example, for calibration purposes using configurable signal routing elements (combined with switching devices for selecting between pulse generator and detector pixel inputs). For simplicity, Figure 2 shows only a single pulse generator 39; however, it should be understood that this circuitry can also be replicated for each pixel and integrated into each pixel.

[0060] The electrical pulses generated by the pulse generator (e.g., each) simulate the electrical detection results of photon interaction events, such as the electrical detection results received from a detector array by a conventional image pixel.

[0061] Pulses can be provided at a predetermined frequency (e.g., a fixed or configurable pulse frequency), or equivalently, the time between subsequent pulses can be fixed or configurable, such that the pulse timing is known and / or precisely controlled. For example, pulse generator 39 can generate charge pulses at a rate of 1 MHz (e.g., in the range of 100 kHz to 10 MHz, or, for example, in the range of 500 kHz to 2 MHz).

[0062] The pulse generator circuit can be used to perform threshold scanning and / or input pulse amplitude scanning, wherein, for example, the output of one or more dummy pixels (e.g., the count value of events that have passed a threshold (or, for example, the count value for multiple thresholds when multiple counting channels / multiple energy bins are implemented)) is observed via an output connection to processor 50, while the dummy pixels receive the generated pulses from the pulse generator circuit as their input.

[0063] The processor 50 may be adapted to control the pulse generator circuit 39 in order to, for example, activate a pulse sequence and control the characteristics of the pulses (e.g., pulse frequency and / or amplitude).

[0064] In a particularly simple embodiment, the pulse generator circuit generates pulses of fixed amplitude at a fixed frequency when the device is active. In a more complex embodiment, the pulse train generated by the pulse generator circuit can cycle through different pulse amplitudes. This cycle can be pre-configured and fixed in the hardware design, so that no active control is required. Alternatively, the pulse generator circuit can still be conveniently controlled by a processor.

[0065] The pulse generator circuit may include a chopper current source or a switched capacitor, a combination thereof, or other suitable devices.

[0066] The pulse generator circuit may include a chopping current source, for example, such that a current having a predetermined and known current value is injected into the input node of the analog front end of the dummy pixel when activated (e.g., by a switching device, for example at a predetermined frequency).

[0067] The pulse generator circuit may include a switched voltage source. For example, the pulse generator circuit may be adapted to receive or generate a predetermined voltage, which in operation may be used to charge a capacitor so that the charge accumulated by the capacitor can be injected into the input node of the dummy pixel.

[0068] What may be particularly advantageous is the use of an arrangement in the pulse generator circuit that relies on a switched voltage source to charge the capacitor, such that the current generated by the charge stored in the capacitor is injected into the input of the dummy pixel (e.g., into the AFE, such as into the preamplifier and / or pulse shaper). The supply voltage received to charge the capacitor is generally insensitive to flicker noise, while the switched current source may result in a current pulse affected by flicker noise. Since the pulse is used to measure the flicker noise effect, it may be important to provide a robust, accurate, and reproducible input pulse to the dummy pixel. For example, the switched current source may typically rely on an implementation using one or more MOSFET elements (i.e., significantly sensitive to flicker noise). By controlling the specific amount of charge injected into the dummy pixel in the pulse using a capacitor, the exact capacitance value of the capacitor is the primary source of uncertainty (or essentially the only source of uncertainty). However, the uncertainty of the capacitance value is not particularly problematic. It is possible to determine that the uncertainty of the capacitance value relative to its design specifications (e.g., due to variations in CMOS manufacturing processes) is within an uncertainty margin of approximately 10% or less, and importantly, the capacitance value will remain substantially constant over time (i.e., will not drift significantly over time).

[0069] Furthermore, calibration of the capacitance value (or more generally, the charge generated by the pulse generator) can be performed, for example, after the device is manufactured. Specifically, if at least one image pixel (e.g., each pixel in the pixel family (dummy pixel and image pixel)) includes a pulse generator structure that is the same as (or substantially similar to) the pulse generator structure used for the dummy pixel, a reference X-ray energy can be used in the calibration. Such a reference energy can be obtained by shielding (one or more) image pixels using, for example, a lead (Pb) filter in a reference X-ray beam and tuning the pixels in response to the K-edge energy (e.g., 88 keV) of the filter (here, Pb). It should be understood that this is merely an illustrative example, and other filters or devices for generating accurately known photon energies can be used. The threshold level of the image pixel (e.g., the threshold level of at least one comparator of the pixel's threshold detector 42) can be adjusted, for example, as a method for calibrating the device to a reference energy level (88 keV in this example) for homogeneous and quantitative image responses.

[0070] The pulse generator circuitry for the image pixel can be controlled to achieve the same or substantially the same output. For example, after reading out the output of the image pixel (e.g., the value generated by the threshold detector 42), the pulse generator circuitry can be connected to the input of the same image pixel (and the connection to the detector pixel is turned off, and / or the detector pixel is de-exposed to photons of the reference energy), and the pulse generator configuration can be tuned (e.g., adjusting the input voltage supplied to the capacitor) to generate the same image pixel output (e.g., substantially the same value). Using this configuration of the pulse generator, an equivalent of the reference energy level is obtained. Since the same pulse generator circuitry (e.g., routed from the same capacitor on the ASIC chip or provided by the same pulse generator circuitry) is used for both the dummy pixel and the image pixel, it is clear that any variation between the capacitor used to determine the reference setting of the pulse generator for the image pixel and the capacitor used to generate the test electrical pulse for the dummy pixel is completely avoided; that is, the control settings of the pulse generator for the dummy pixel can be made the same as the control settings of the pulse generator circuitry for the image pixel corresponding to the actual photon interaction event to achieve substantially the same simulated photon energy. If each pixel has its own dedicated pulse generator circuitry integrated therein, it is still reasonable to assume that the variation between different capacitors on the same ASIC is minimal (e.g., for ASIC components on the same chip, this variation is small). Therefore, it is possible to use calibrated pulse generator settings (selecting the amplitude of the test current pulse) on dummy pixels to generate a reference test pulse corresponding to a calibrated reference energy (e.g., 88 keV).

[0071] For example, the operator can select to perform calibration, for instance, via the user interface of the console or the detector's control system. This can be performed using a detector assembled in the system for its intended purpose (e.g., a CT system), but can also be performed alternatively (e.g., during / after manufacturing) in a dedicated calibration setting. In the former case, radiation incident on the detector array module can be provided by an X-ray tube designed for system imaging. To implement a well-defined reference spectrum, a K-edge filter or other spectral filter with known characteristics can be placed in the beam, for example, inserted between the radiation-sensitive side of the detector array module and the X-ray tube. The filter allows for the acquisition of photons with a known and robust predetermined photon energy (i.e., the actual X-ray photon energy is precisely known by the specific K-edge of the filter material), and the ASIC image pixel output can be observed in response to such photons with a known and robust predetermined photon energy. For example, for each image pixel, the detector pixel output value can be determined in response to the known photon energy (e.g., ADC quantization, filtering, and / or pre-tuned pulse amplitude and / or observation count for predetermined settings of one or more thresholds). The above operations can be repeated for different energies (e.g., using different edge filters) to characterize this relationship in more detail. It is possible to adjust threshold settings (e.g., threshold table of the image pixel threshold detector 42 comparator) and / or correct data obtained from the detector to homogenize the response (e.g., per-pixel output correction map) using calibrations related to photon energy and ASIC pixel output.

[0072] In the absence of incident radiation (e.g., when the tube is off) and when the pulse generator circuitry is switched to serve as an alternative input signal source for the image pixels, the pulse generator circuitry can be tuned to achieve the same pixel output. Thus, a pulse generator setting that accurately simulates the reference input (i.e., the reference photon energy) is obtained. This can include tuning the supply voltage to charge the capacitors in the pulse generator circuitry, and / or optimizing the charging time (e.g., shortening the time the capacitors are charged before releasing charge, for example, by adjusting the frequency of the pulse cycle, typically results in lower charge accumulation). When differences between pulse generator circuitry across the ASIC chip (e.g., differences in capacitance values) can be assumed to be negligible (which would generally be an acceptable assumption), the same pulse generator settings can be applied to dummy pixels to simulate the same reference energy incident photon event. If the image pixels and dummy pixels can be controlled to receive input from the same pulse generator circuitry, for example, by using controlled signal routing (e.g., a signal routing switch), any variation in capacitance between the image pixels and dummy pixels can be eliminated.

[0073] The pulse generator circuit may also include a galvanometer for measuring test pulses injected into the signal processing chain for calibration. Calibration of the pulse generator circuit to simulate a specific reference energy and / or to verify the generated current pulses can be performed directly by measuring the injected current (or its integral, i.e., the injected charge), regardless of uncertainties in the generated current (e.g., due to uncertainties in capacitance values) and / or instabilities (e.g., fluctuations in the generated current, e.g., when using a chopper current source).

[0074] Advantageously, dummy pixels can be used to more accurately determine one or more thresholds of charge packets corresponding to a predetermined energy (e.g., 75 keV, but not limited thereto), and / or to perform operations targeting a predetermined threshold (e.g., a fixed threshold). fT (e.g., 75keV, but not limited thereto) input pulse scan.

[0075] Therefore, further signal processing in the image chain, for example, implemented by a processor, may include: receiving extracted pulse amplitude and / or counting information (e.g., as a function of time, such as as a time series or time-varying signal), and processing said information to, for example, correct the information received from the image pixels (e.g., perform a correction to the observation count (CoC)). For example, the original image information provided by the ASIC can be corrected by taking into account dummy pixel information (e.g., combining a conversion factor between amplitude and observation count (OC) (e.g., at a photon energy of 75 keV).

[0076] For example, a conversion factor can be empirically determined, for instance, based on measurements of the integral spectrum and / or simulations of the detector design, thereby converting the detected amplitude and / or threshold variations (discussed below) into relative changes in the observation count (OC) for application to image pixel counts. Thus, pulse amplitude information can be used to determine changes in the observation count by applying a conversion factor. While a conversion factor is cited, it should be understood that if simulations or experiments show a significant deviation from perfectly linear behavior, the relationship between the amplitude variation of the observation and the corresponding observation photon count (OC) can also be expressed (e.g., empirically determined or simulated) by a more complete description of the conversion function (e.g., a higher-order polynomial, interpolation in a lookup table of values, and / or any other suitable way of representing the functional relationship). Nevertheless, given that this relationship is between relative quantities (variations in amplitude and OC) and assuming that these variations have a relatively limited range in practical use, it can be assumed that the conversion is substantially linear (proportional).

[0077] For example, for a 75 keV test energy example, the relative change in observed counts can be approximately 6 to 7 times larger than the relative amplitude change in observations. For instance, for a typical PhC detector design (e.g., a CZT photon counting spectrometer for CT), the conversion factor can be in the range of 4 to 7, for example, in the range of 5 to 8, for example, in the range of 6 to 7, and is not limited to these examples. The conversion factor (or function) can typically depend on the photon energy (and / or spectrum). Note that it is possible to control the pulse generator circuitry to accurately obtain the reference test energy, i.e., to simulate the reference photon energy with good fidelity, where the reference test energy is insensitive to or has negligible sensitivity to time-varying effects (e.g., flicker noise in components of the pulse generator circuitry itself). However, counts observed through image pixels will typically be generated in response to a multicolor spectrum, which may necessitate consideration of the conversion function.

[0078] exist Figure 4 The illustrative example shown illustrates an input pulse scan (ipS), where a pulse signal provided by a pulse generator circuit varies in amplitude across different pulses P in the pulse train, for example, by gradually increasing (or decreasing; but not limited to monotonic and / or constant step amplitude variations). An illustrative fixed threshold is also shown in this example. fT At low pulse amplitudes, the peak value of the pulse is insufficient to exceed the fixed threshold level, so that the number of counted pulses (observed events) remains zero according to the threshold setting.

[0079] However, when the nominal pulse amplitude is below the threshold, but the sum of the Gaussian noise amplitudes is sufficient to exceed the threshold, a non-zero number of events will be recorded, and the number of observed events will correspond to the measurement period when the input pulse amplitude exceeds the threshold level. Multiply by pulse frequency ,Right now, That is, the response will saturate to the theoretical maximum value.

[0080] This simple example illustrates that, for a fixed threshold setting, the number of observations observed by a dummy pixel in response to a test pulse (as a measure of the test pulse amplitude, e.g., the applied voltage step size) is... multiples of The function representing the input pulse amplitude will have a Gaussian error function. erf The shape of (e.g., ignoring x and / or y scaling and / or offset) (e.g., as Figure 5As shown, the shape of the complementary error function mirrored along the y-axis is given when the noise is assumed to be Gaussian distributed or can be sufficiently approximated by it (a reasonable assumption in the current context). Based on the analysis of the observation counts as a function of pulse amplitude (or, as a function of time when the amplitude monotonically increases or decreases during the scan), the inflection point TP (e.g., a crossover or inflection point, e.g., where the second derivative changes sign, and / or where the curve changes from concave to convex, or vice versa) can be readily determined. For example, a point or a sufficiently approximation thereof can be detected when the observation count reaches half of the theoretical saturation value (the maximum expected count based on the number of pulses generated per time frame). The x-coordinate in this representation (i.e., the pulse amplitude corresponding to the inflection point) can be considered a measure of the true pulse amplitude in the absence of noise. The circuitry of the dummy pixel can be adapted to determine the inflection point and output the value (the amplitude of the input pulse generated by the pulse generator circuit where the TP is located) to the image chain, and / or the processor in the image chain can be adapted to determine the inflection point based on the raw signal output of the dummy pixel (e.g., by receiving a count of events exceeding a fixed threshold set for each pulse amplitude of the pulse generated by the pulse generator).

[0081] When flicker noise affects the ASIC, it causes the amplitude to change over time. The pulse scanning method allows for the detection of this amplitude change, enabling the threshold level to be reset to the detected TP amplitude.

[0082] In an alternative method, the pulse amplitude of the pulses generated by the pulse generator circuit can be kept constant, and the threshold setting can be adjusted for different pulses in the pulse train. Therefore, threshold scanning can be performed. In this case, such as... Figure 6 As shown, a fixed-amplitude current pulse P is injected into one or more dummy pixels, while simultaneously (e.g., using trimming settings of the dummy pixels' threshold detector) scanning a threshold level within a certain range. vT (The threshold for change). At a small threshold level, record all injected pulses, and the number of observation counts should be equal to... Once the selected threshold level reaches the nominal pulse amplitude, OC decreases due to noise, and when the threshold level is higher than the pulse amplitude plus the maximum noise amplitude (randomly speaking, a large standard deviation of noise above the pulse amplitude), the number of observation counts drops to essentially zero. The waveform of the observation count OC as a function of the threshold level when Gaussian noise is assumed (see...) Figure 7 ) follows the complementary Gaussian error function ( erfcThe shape of the pulse is determined, and again, the x-coordinate of the inflection point (TP) is identified as the threshold level corresponding to the nominal pulse amplitude without noise, which can be reported to the image chain (and / or determined by the image chain). Therefore, any amplitude variation over time caused by flicker noise can be observed and corrected.

[0083] The device can repeatedly perform measurements to monitor changes in the device's output caused by flicker noise over time. For example, the scanning method discussed above can be repeatedly performed during the use of the device (see...). Figure 4 and Figure 6 The inflection point can be repeatedly determined to monitor changes caused by noise over time. The collected TP estimates can also be averaged over time (e.g., by averaging over 100 views (this is an example, but not limited to)) to remove high-frequency variations, for example, to avoid or reduce sensitivity to instantaneous changes in flicker noise and / or measurement noise and / or improve the robustness of the method. If the inflection point (its x-coordinate) changes over time, the processor (i.e., the image chain) can adjust the observation count (OC) (e.g., the raw image pixel count) related to the image content based on this detected amplitude change in one or more dummy pixels. Based on the change in pulse amplitude over time, the image chain is able to adjust the OC in each frame to correct for flicker noise effects using, for example, a known conversion factor between a pulse amplitude shift of a 75 keV simulated (single-energy) photon and an OC shift (for the typically multicolor spectrum used in actual imaging).

[0084] For example, as discussed above, the test pulse can reach a turning point in the detected DAC value (controlling the pulse generator circuitry), which corresponds to a specific observation count value (e.g., a known correspondence determined through calibration / simulation of a predetermined target photon energy, e.g., 75 keV). Therefore, the observation count (image pixel data) in each image frame can be adjusted based on this OC value corresponding to the detected amplitude shift. The DAC value at the turning point can be reported to the image chain as a measure of the observed pulse amplitude every 100th frame (but not limited to) to monitor changes in pulse amplitude over time, based on which the corresponding OC shift to be applied to correct the image data can be determined. The collected TP data can also be averaged over time (e.g., by averaging over 100 pulses (this is an example, but not limited to)) to remove high-frequency variations, thereby, for example, improving the robustness of the method. Based on the change in pulse amplitude over time, the image chain can adjust the observation count values ​​(of image pixels) accordingly (e.g., subtract the estimated noise-related counts from the observation values ​​estimated from the output of the monitored dummy pixels).

[0085] Advantageously, threshold scanning and / or pulse amplitude scanning can be performed during image data acquisition because dummy pixels will not be affected by X-ray photons.

[0086] The device can also repeatedly perform measurements to monitor changes in flicker noise in the device's output over time by setting one or more thresholds of the dummy pixels to levels detected according to the methods discussed above (e.g., setting a threshold scan inflection point (TP) for fixing the amplitude of the input test pulse). In response to at least one input test pulse (e.g., having the same amplitude as the TP used to determine the threshold (threshold scan) via the response curve), observation counts of one or more dummy pixels at this threshold setting can be collected during use of the device (e.g., shortly before, shortly after, or during image acquisition). If the observation count is significantly higher or lower than expected (e.g., causing the measurement result of the dummy pixel to correspond to along...), the device will monitor for changes in flicker noise in the device's output over time. Figure 7 If the illustrative curve in the diagram shifts to the right and down or to the left and up at a point, it can be assumed that the flicker noise level has changed relative to the initial threshold determination (e.g., by the threshold scan discussed above). Therefore, it is possible to correspondingly correct the image pixel data collected simultaneously or approximately at the same time point, for example, (e.g., before / during the image reconstruction process) using OC values ​​obtained from (one or more) dummy pixels in the image chain to correct the reported (image pixel) count (CoC).

[0087] The device can also repeatedly execute tests on a fixed threshold using a constant pulse amplitude generated from the test pulses. fT The measurement of the observation count (OC) is determined by the inflection point (TP) determined by the input pulse sequence (ipS) discussed above (e.g., corresponding to the inflection point (TP) determined by the input pulse sequence (ipS) discussed above) (e.g., see...). Figure 4 and Figure 5 If the observed count is significantly higher or lower than the expected value (e.g., detected by applying a predetermined threshold to the difference), then the current measurement corresponds to along... Figure 5 The illustrative curve in the diagram shifts to the right and up, or to the left and down (depending on the sign of the difference). Therefore, such a significant change can indicate that the flicker noise level has changed relative to the initial conditions when determining the pulse amplitude based on the inflection point of the iPS. This also allows for monitoring changes over time. Consequently, it is possible to adjust the count (CoC) correction applied to the image pixel data to compensate for such detected changes, and / or trigger new, more comprehensive tests for accurately determining the flicker noise level (e.g., via the fT method and / or iPS method discussed above).

[0088] While a single dummy pixel can be used to estimate flicker noise, it should be understood that using multiple dummy pixels allows for a more accurate determination of flicker noise, for example, by averaging the results obtained from different dummy pixels (and / or by estimating the results based on the average output from different dummy pixels). This allows for the identification of low-frequency drifts common to all pixels on the ASIC, particularly drifts that commonly affect dummy pixels (and, by extension, can also be assumed to affect image pixels).

[0089] Flicker noise exhibits low-frequency drift affecting all pixels of the same ASIC, and this common drift often leads to image artifacts (e.g., in CT reconstruction). However, (irrelevant) single-pixel flicker noise may also exist, which typically results in different types of artifacts. Automatic zeroing compensation methods are known in the art, capable of compensating for the aforementioned situations at least to some extent. Furthermore, flicker noise-optimized CMOS designs are known in the art, designed to reduce the impact of flicker noise; for example, such designs may use MOSFETs larger than those typically used. Since individual pixel variability can be reduced using known methods, a number of dummy pixels may be sufficient, if needed (e.g., the aforementioned automatic zeroing compensation and / or design optimization), to obtain enough data to characterize the associated low-frequency drift, for example, to enable averaging of any local deviations in the response of the dummy pixels. By averaging the acquired data over at least a few (e.g., at least 2, at least 4, at least 8, or at least 16) dummy pixels, (low-frequency) noise from unrelated dummy pixels can be effectively suppressed. It can also eliminate (or sufficiently reduce) high-frequency noise by averaging over, for example, a sufficiently large number of views (time samples) over time, enabling robust and efficient determination of the observation count and / or amplitude of dummy pixels, and accordingly correcting the image pixel signal with good accuracy.

[0090] refer to Figure 8 , Figure 8Illustrative results of the method discussed above are shown, where the OC value is measured at a fixed threshold. At time 0.2 s, the X-ray shield is opened (this can be observed from a sudden surge of OC values ​​82 (grey) reported by multiple image pixels (averaged across multiple pixels), i.e., the ASIC image pixels measure the response to incident radiation at a fixed threshold of 75 keV. Low-frequency fluctuations in the measured X-ray count signal are obtained through a smoothing operation and are shown in the overlapping curve 81 (white) on the original measurement results. Multiple dummy pixels (22 pixels in this example) (here, a single row) are not used for X-ray photon conversion (in this example, X-ray photons are blocked by a lead filter). These dummy pixels collect OC values ​​at the fixed threshold setting in parallel with image acquisition, where an electrical test pulse corresponding to 75 keV is supplied. The fixed threshold for the dummy pixels is set to the inflection point of the threshold scan (performed prior to the illustrated test results), as discussed above (the same amplitude for the 75 keV test pulse). The smoothed result of the average OC across the dummy pixels is shown in graph 83 (dark black). Clearly, this dummy pixel signal is independent of the X-ray beam and therefore exists even when the shackle is still closed (before 0.2 s). It can be clearly observed that the dummy pixel output 83 and the smoothed image pixel output 81 exhibit very similar fluctuations. Since the pulse generator reference result (observed by the dummy pixels) is not affected by the non-ideals of the CZT sensor, these fluctuations must originate from the electronics (i.e., the readout circuitry). Therefore, this pulse generator reference can be used to correct for fluctuations in the X-ray signal (e.g., image pixel signals used for image reconstruction).

[0091] Figure 9 The result of this correction is shown. In this example, the shader is opened at sampling time index 800 (arbitrary unit). The average signal of a set of X-ray pixels (e.g., 300 X-ray pixels in this example) before correcting the OC continuously monitored at a fixed threshold output using this set of dummy pixels is shown for several different acquisitions performed consecutively. Each grayscale value represents a different image acquisition in the series. Reference Figure 10 , Figure 10 The results for the same set of pixels (for different acquisitions) after applying correction based on the observed OC values ​​of dummy pixels are shown. It can be seen that the correction significantly smooths the measured X-ray count traces. For example, in Figure 9 As shown by arrow 91, for one of the multiple acquisitions, a significant offset can be observed near time index 4500. This offset was effectively corrected by applying this method (see [link to documentation]). Figure 10 (where there is no such offset).

[0092] These illustrative results also demonstrate that operating the ASIC according to embodiments of the invention enables multiple pixels to receive electrical test pulses without causing any substantial (obvious) interference. The improved stability resulting from the correction count significantly outweighs any potential disturbances (e.g., interference) caused by the use of pulse generator circuitry in dummy pixels (e.g., due to sufficient isolation between pixels on the same chip).

[0093] refer to Figure 11 In a fourth aspect, the present invention relates to a method 10 for characterizing, reducing and / or correcting flicker noise in a photon counting image detector, for example, according to an embodiment of the first aspect of the invention.

[0094] The method includes, for example, using a pulse generation circuit to generate (11) a plurality of electrical test pulses, and providing the electrical test pulses to at least one dummy pixel of an application-specific integrated circuit of a photon counting image detector.

[0095] The method includes: obtaining (12) a charge packet via a connection to a photon counting detector array module, and providing the charge packet in the form of an electrical input signal to the image pixel circuit of an application-specific integrated circuit.

[0096] The method includes processing (13) a dummy pixel output signal (or multiple signals of multiple dummy pixels) generated by the dummy pixel based on an electrical test pulse by comparing an electrical input pulse or a signal derived therefrom with at least one threshold.

[0097] The method includes processing (14) an image pixel output signal generated by an image pixel based on a charge packet by comparing a charge packet or a signal derived therefrom with at least one threshold.

[0098] In the method according to the embodiment, an electrical test pulse can be generated by a chopper current source and / or by charging a capacitor having a predetermined capacitance (e.g., by charging with a switched voltage source) (and subsequently discharging).

[0099] Although exemplary embodiments are described below, the invention is limited only by the appended claims.

[0100] As used in the claims, the word "comprising" is not limited to the features, elements, or steps described herein, and does not exclude additional features, elements, or steps. Therefore, this specifies the presence of the mentioned features without excluding the further presence or addition of one or more features.

Claims

1. A device (20) for photon counting imaging, the device comprising: A photon counting detector array module (30) includes detector pixels (34) for converting photons into charge packets, and Application-specific integrated circuit (32), comprising: Multiple image pixel circuits (35), operably connected to the corresponding detector pixels (34) of the photon counting detector array module (30), and At least one dummy pixel circuit (36) is not electrically connected to any detector pixel of the photon counting detector array module. The device is characterized in that it further includes: At least one pulse generation circuit (39) is connected to the at least one dummy pixel circuit for generating a plurality of electrical test pulses. A processor (50) adapted to analyze the output signal of the at least one dummy pixel circuit (36), which is generated in response to the plurality of electrical test pulses, in order to determine the following metrics, and thereby process and correct the output of the image pixel circuit (35) and / or adjust the settings of the circuit (32) by taking the metrics into account, the metrics indicating at least one parameter characteristic of at least one of the operation of the application-specific integrated circuit (32) and the operation of the detector array module (30) drifting, changing and / or deviating.

2. The device according to claim 1, wherein, Each of the plurality of image pixel circuits is adapted to compare an electrical input pulse from or derived therefrom of the corresponding detector pixel with at least one threshold, and wherein the at least one dummy pixel circuit is adapted to compare the plurality of electrical test pulses or derived therefrom with at least one threshold. The processor (50) is adapted to analyze the output of the at least one dummy pixel circuit (36) generated in response to the plurality of electrical test pulses in order to determine the metric to quantify the difference between the at least one threshold and the corresponding test pulse amplitude, wherein the output of the at least one dummy pixel circuit is a function of the pulse amplitude of the electrical test pulse and / or a function of the setting of the at least one threshold of the dummy pixel circuit, and The processor (50) is adapted to process and correct the output of the image pixel circuit (35) and / or adjust the setting of the at least one threshold of the image pixel circuit by taking into account the metric that quantifies the difference.

3. The device according to claim 2, wherein, The pulse generation circuit (39) is adapted to generate the electrical test pulse as an amplitude scan sequence covering a plurality of electrical test pulses with different pulse amplitudes, and the processor (50) is adapted to analyze the output of the at least one dummy pixel circuit (36), the output of the at least one dummy pixel circuit being generated in response to the amplitude scan sequence and with the at least one threshold being constant, the output of the at least one dummy pixel circuit being a function of the plurality of different pulse amplitudes.

4. The device according to claim 2 or claim 3, wherein, The pulse generation circuit (39) is adapted to generate the electrical test pulses into a fixed amplitude sequence of electrical test pulses having the same amplitude corresponding to a predetermined photon energy, and wherein the processor (50) is adapted to analyze the output of the at least one dummy pixel circuit (36), the output of the at least one dummy pixel circuit being generated in response to the fixed amplitude sequence and in the event that the value of the at least one threshold is varied over a plurality of different threshold evaluation values, the output of the at least one dummy pixel circuit being a function of the plurality of different thresholds.

5. The device according to claim 3 or claim 4, wherein, The processor (50) is adapted to determine the measure of quantifying the difference by determining the pulse amplitude and / or the threshold corresponding to the inflection point (TP) between the low and high states of the output, the low state corresponding to substantially zero pulses exceeding the threshold, and the high state corresponding to substantially all pulses exceeding the threshold.

6. The device according to claim 5, wherein, The processor (50) is adapted to correct the output signal of the image pixel circuit (35), which includes the observation count value, by taking into account the pulse amplitude and / or the threshold corresponding to the inflection point (TP), wherein at least one conversion factor is applied to correlate the change in the pulse amplitude and / or the threshold of the inflection point with the corresponding change in the observation count value to obtain the correction result for a reference integrated X-ray spectrum.

7. The device according to claim 5, wherein, The processor (50) is adapted to: Determine the threshold corresponding to the transition point (TP) between the low and high states of the output, where the low state corresponds to substantially zero pulses exceeding the threshold and the high state corresponds to substantially all pulses exceeding the threshold. Control the at least one dummy pixel circuit to set the at least one threshold of the at least one dummy pixel circuit to a determined threshold corresponding to the inflection point, and Subsequent changes in the output of the at least one dummy pixel circuit are monitored in response to an additional electrical test pulse of the amplitude corresponding to the predetermined photon energy, wherein the subsequent changes represent a measure of the difference in the output of the image pixel circuit (35) for correcting the difference.

8. The device according to any one of the preceding claims, wherein, The processor (50) is adapted to repeatedly analyze the output of the at least one dummy pixel circuit and determine the measure used to quantify the drift, the change and / or the deviation in order to monitor the evolution of the at least one parameter over time.

9. The device according to any one of the preceding claims, wherein, The pulse generation circuit (39) includes: a chopper current source for generating the plurality of electrical test pulses, and / or a capacitor with a predetermined capacitance connected to a switching voltage source.

10. The device according to any one of claims 2-9, wherein, The pulse generation circuit (39) is adapted to be connected to at least one of the image pixel circuits in the image pixel circuit (35) via a network of controllable switches and / or controllable signal routers, so as to route the electrical test pulse generated by the pulse generator circuit as the electrical input pulse to the image pixel circuit for energy calibration purposes in a first switching setting, and to route the electrical input pulse from the detector pixel to the image pixel circuit for image acquisition purposes in a second switching setting.

11. The device according to any one of the preceding claims, comprising: The dedicated pulse generation circuit is essentially the same for each dummy pixel circuit and for each image pixel circuit, wherein the dedicated pulse generator circuit is adapted to exclusively provide the electrical test pulse to the corresponding dummy pixel circuit and / or image pixel circuit.

12. A diagnostic imaging system (100) comprising a photon counting image detector according to the preceding claim.

13. The application-specific integrated circuit (32) according to claim 1, wherein, Multiple image pixel circuits (35) are configured to be operatively connected to the corresponding detector pixels (34) of the photon counting detector array module (30), wherein no external connection is provided so that the dummy pixel circuit (36) cannot be electrically connected to any detector pixel of the photon counting detector array module.

14. A method (10) for characterizing, reducing, and / or correcting drift, variation, and / or deviation of at least one parameter characteristic of at least one of the operation of an application-specific integrated circuit (32) and the operation of a detector array module (30) in a photon counting image detector, the method comprising: (11) Generate (11) multiple electrical test pulses and provide the electrical test pulses to at least one dummy pixel circuit of the application-specific integrated circuit of the photon counting image detector, wherein the at least one dummy pixel circuit generates an output signal based on the electrical test pulses, and wherein the at least one dummy pixel circuit is not electrically connected to any detector pixel of the photon counting detector array module. The electrical input pulses originating from the corresponding detector pixels are provided (12) to the corresponding image pixel circuits of the application-specific integrated circuit. Analyze (13) the output signal of the at least one dummy pixel circuit to determine a measure indicating the drift, variation, and / or deviation of at least one parameter, the output signal of the at least one dummy pixel circuit being generated in response to the plurality of electrical test pulses, and The image pixel circuit output signal is corrected (14) and / or the settings of the application-specific integrated circuit are adjusted by taking the metric into account.

15. A computer program product for executing the method of claim 14 when run by a processor.

Images