Recalibration of a radiation detector

Abstract

The invention relates to a method for calibrating a radiation detector with a plurality of pixels. In this method, a radiation spectrum from a radiation source is passed through the radiation detector, with a plurality of energy-dependent count rates for different signal thresholds of an evaluation unit of the radiation detector being measured for each pixel of the radiation detector. Updated signal threshold-count pairs are then determined based on the different energy thresholds and their corresponding count rates. Furthermore, pixel-specific shift values ​​for the signal thresholds are determined by comparing the updated signal threshold-count pairs with previous pixel-specific reference signal threshold-count pairs from a prior calibration of the radiation detector.Furthermore, recalibrated signal threshold-signal energy value pairs are determined pixel by pixel by adding the respective shift value to the signal threshold values ​​of the respective reference signal threshold-signal energy value pairs of the respective pixel from the previous calibration. The radiation detector is then recalibrated using the recalibrated signal threshold-signal energy value pairs. The invention further relates to a recalibration device, a radiation detector, and an imaging system.

Description

[0001] The present invention relates to a method for recalibrating a radiation detector having a plurality of pixels. Furthermore, the invention relates to a calibration device. Furthermore, the invention relates to a radiation detector. Furthermore, the invention relates to an imaging system.

[0002] Modern imaging techniques often generate two- or three-dimensional image data that can be used to visualize an object under examination and for other applications. To obtain image information from an object under examination, systems based on the detection of electromagnetic waves that have interacted with the object under examination are often used.

[0003] One way to acquire image data from the interior of such an object under examination is based on X-ray imaging technology. An X-ray system used for this purpose comprises an X-ray source for generating X-rays and an X-ray detector for detecting the X-rays. The X-ray detector comprises a detection unit that generates a detection signal for X-rays impinging on the detection unit, and an evaluation unit that determines a set of count rates for X-rays impinging on the detection unit based on the detection signal.

[0004] In order to acquire three-dimensional data from an examination subject, a computed tomography system (CT system for short) is frequently used to acquire projection measurement data from the examination subject. In CT systems, a combination of an X-ray source and an opposite X-ray detector, arranged on a gantry, usually rotates around a measurement space in which the examination subject (hereinafter referred to as the patient without loss of generality) is located. The center of rotation (also called the "isocenter") coincides with a so-called system axis z. In one or more rotations, the patient is irradiated with X-rays from the X-ray source, with projection measurement data or X-ray projection data being acquired with the help of the opposite X-ray detector.

[0005] The generated projection measurement data depends particularly on the design of the X-ray detector. X-ray detectors typically have a plurality of detection units, usually arranged in the form of a regular pixel array. The detection units each generate a detection signal for each X-ray incident on the detection units, which is analyzed at specific times with regard to the intensity and spectral distribution of the X-ray radiation in order to obtain information about the object under examination and generate projection measurement data.

[0006] So-called quantum-counting X-ray detectors are now being used to detect projection measurement data. In such quantum-counting or photon-counting X-ray detectors, the detection signal for X-ray radiation is analyzed with regard to the intensity and spectral distribution of the X-ray radiation in the form of count rates. The count rates are provided as output data of a so-called detector channel, each of which is assigned to a detection unit. In quantum-counting or photon-counting detectors with multiple energy thresholds, each detector channel usually generates a set of count rates per projection based on the respective detection signal of the detection unit. The set of count rates can include count rates for several different energy thresholds, especially those checked simultaneously.The energy thresholds and the number of energy thresholds, each of which is assigned an energy threshold, are usually specified as signal analysis parameters for capturing the projection.

[0007] The control and signal analysis in an evaluation unit of such an X-ray detector is carried out by externally supplied register values, which are converted into control voltages via digital-to-analog converters (DACs for short). In this context, a digital setting is also referred to below as a DAC value or signal threshold value, which specifies the control function by naming the detector parameter. The DAC values ​​correspond, for example, to control voltages of the X-ray detector, which set electronic thresholds of comparators in the signal evaluation unit. The electronic thresholds of the comparators, for example electrical voltage values, in turn correspond to energy thresholds, with the help of which the energy of incoming X-ray quanta can be measured and differentiated.The energy thresholds define a range of energy values ​​in which a comparator detects X-ray quanta and thus counts the X-ray quantum. X-ray detectors usually have several electronic thresholds per channel, which are controlled with different DAC values ​​or signal thresholds. In order to be able to measure the energy of the incident X-ray quanta absolutely, an absolute energy scale must be calibrated. This means that a relationship must be determined between the electronic signal thresholds of the comparators and the corresponding energy thresholds for incident X-ray quanta, from which a signal energy value can be determined. Thus, when measuring X-ray quanta, the measured signal threshold value is assigned to a signal energy value based on knowledge of the aforementioned relationship and on the basis of knowledge of a measured signal threshold value.

[0008] Typically, calibration, also known as tuning, is performed during the initial tuning of an X-ray detector or a CT system with such an X-ray detector during production. Furthermore, an X-ray detector must also be recalibrated if necessary to improve accuracy after delivery to the customer.

[0009] However, the settings of the energy thresholds or the electronic thresholds corresponding to the energy thresholds in fixed steps, the DAC values ​​or signal thresholds, are only possible with a certain degree of granularity. For example, 1 DAC can correspond to an energy of 0.5 keV. Therefore, the electronic thresholds cannot be precisely adjusted; instead, binning or rounding to whole integer values ​​takes place before the final tables are created. If the sensor signals change slightly over time, it is necessary to adjust or optimize this setting.

[0010] If image quality artifacts occur that indicate a spectral change in the X-ray detector, a complete retuning is traditionally performed. However, such a process is very time-consuming, requiring, for example, a full workday during which a technician is working on the system and the system is unavailable to a radiologist.

[0011] Furthermore, noise effects often make it difficult to accurately determine the characteristic structures of a characteristic X-ray spectrum. Due to these noise effects, a certain number of pixels are traditionally combined for spectral fine-tuning, since only combining and averaging over a large number of pixels makes the spectral structures recognizable.

[0012] During such a retuning, the energy thresholds or the electronic thresholds assigned to these energy thresholds are completely remeasured and set to a fixed value. However, this step introduces all measurement and evaluation noise, including rounding / sampling errors, into the tables. Therefore, all subsequent tuning tables become invalid and must be adjusted accordingly.

[0013] The document DE 10 2013 200 021 A1 discloses a method for calibrating a counting, digital X-ray detector, wherein a threshold scan is carried out for a defined X-ray spectrum and wherein individual correction thresholds are determined by means of the counting rates measured by the pixel elements.

[0014] The document US 2024 / 0 230 934 A1 discloses a method for calibrating a detector comprising recording an energy spectrum obtained from a scan with an X-ray tube as the radiation source, wherein a trained neural network is used to determine calibration parameters.

[0015] The object of the invention is to provide a method for calibrating radiation detectors, in particular X-ray detectors in CT systems, which is less complex and in particular less time-consuming than the conventional methods for calibrating a radiation detector.

[0016] This object is achieved by means of a method for recalibrating a radiation detector according to claim 1, a calibration device according to claim 11, a radiation detector according to claim 12 and an imaging system according to claim 13.

[0017] In the method according to the invention for recalibrating a radiation detector with a plurality of pixels, a radiation spectrum of a radiation source, preferably an X-ray spectrum, particularly preferably a characteristic X-ray spectrum of an X-ray source, is first detected by the radiation detector. The radiation spectrum represents the energy distribution of the radiation emitted by the radiation source. It is important that the radiation spectrum used for recalibration is as similar as possible to an older radiation spectrum measured during a previous calibration or recalibration, and that this older radiation spectrum and its associated calibration or recalibration are used as a reference for the current recalibration. The high level of similarity can be achieved by identical control of the radiation source during the reference measurement and the current recalibration.In this context, a radiation detector is understood to be a device that detects electromagnetic waves or particle radiation. The radiation source used for recalibration should therefore be the same as the radiation source used for the previous calibration.

[0018] During the recalibration measurement, a plurality of energy-dependent count rates for different signal thresholds or electronic thresholds of an evaluation unit of the radiation detector are measured for each pixel of the radiation detector, and updated signal threshold-count pairs are determined based on the measured count rates and the signal thresholds or DAC values ​​assigned to them.

[0019] These DAC values, or signal thresholds, are thresholds of electronic thresholds of digital-to-analog converters that generate digital values ​​based on sensor signals. By comparing the updated signal threshold-count pairs with the reference signal threshold-count pairs of a previous reference measurement, which preferably includes a full calibration, a shift in the signal thresholds of the updated signal threshold-count pairs relative to the older signal thresholds of the reference signal threshold-count pairs is calculated.

[0020] In addition to the signal threshold-count value relationship characterized by the reference signal threshold-count value pairs, the previous reference measurement is also associated with a signal energy value-signal threshold value relationship with a reference signal energy value and a reference signal threshold value. The latter relationship can be derived based on the signal threshold-count value relationship and knowledge of the characteristic properties of the radiation spectrum underlying the reference measurement.

[0021] As explained in more detail later, the shift in the signal thresholds is preferably initially determined as a continuous value for correcting the signal energy value-signal threshold relationship of the reference measurement. The corrected signal energy value-signal threshold relationship thus determined is rounded to discrete DAC values ​​or signal thresholds only upon completion of the recalibration.

[0022] Thus, pixel-by-pixel, pixel-specific shift values ​​for the signal thresholds are determined by comparing the updated pixel-specific signal threshold-count pairs with pixel-specific reference signal threshold-count pairs determined at an earlier time from a previous calibration of the radiation detector.

[0023] Finally, recalibrated signal thresholds are determined by pixel-by-pixel addition of the respective pixel-specific shift value to the signal thresholds of the respective pixel-specific assignments between the reference signal threshold and reference signal energy value of the respective pixel. In other words, recalibrated signal threshold-signal energy value pairs are determined by pixel-by-pixel addition of the respective pixel-specific shift value to the signal thresholds of the respective pixel-specific reference signal threshold-signal energy value pairs of the respective pixel of the previous calibration.

[0024] The radiation detector is then recalibrated using the recalibrated signal threshold-signal energy value pairs. During the recalibration process, the signal threshold-signal energy value pairs stored in the radiation detector's data memory are modified based on the determined signal threshold-signal energy value pairs and made available for correction of future measurements.

[0025] Advantageously, the reference signal thresholds of the pairs of signal thresholds and signal energy values ​​determined during the original calibration can be corrected pixel by pixel without having to redetermine the signal energy values. Since the determined shift in the signal thresholds is usually very small, the respective signal threshold does not change for most pixels due to the rounding of the signal thresholds to integer values. Only for pixels where the shift causes a change in a rounded signal threshold does the assignment between the rounded signal threshold and the signal energy value need to be changed. These are usually pixels where a non-rounded signal threshold is very close to a rounding threshold.

[0026] This reduces the effort required for recalibration compared to a complete recalibration, since the signal energy values ​​do not have to be remeasured and rounded signal thresholds, i.e. DAC values, do not have to be changed for each pixel.

[0027] Furthermore, recalibration can be performed without a detailed analysis of the acquired count rate spectrum with regard to prominent spectral lines and their associated energy values. Such an analysis is usually hampered by strong noise effects, so that conventional recalibration is associated with increased uncertainty in determining the individual energy thresholds or the signal threshold-signal energy value pairs. In a conventional calibration, several pixels are usually combined, and the energy thresholds or signal threshold-signal energy value pairs are determined together for them in order to reduce the aforementioned noise effects and determine the prominent spectral lines with some degree of reliability.Advantageously, in contrast to conventional recalibration, the inventive procedure allows for pixel-by-pixel correction, and the aforementioned noise effects can be suppressed, since the aforementioned spectral information is not required to perform the recalibration. Recalibration can therefore be performed with increased detector resolution compared to the conventional approach. Due to the increased resolution of the precise detection of X-rays, the overall image quality of a radiation detector recalibrated using the inventive method, in particular of an X-ray detector of this type, can be improved.

[0028] The recalibration device according to the invention for a radiation detector has a data reception interface for detecting a radiation spectrum of a radiation source, preferably an X-ray spectrum of an X-ray source, of the radiation detector. For each pixel of the radiation detector, a plurality of energy-dependent count rates for different signal thresholds are received from an evaluation unit of the radiation detector, and updated signal threshold-count pairs are determined based on the different signal thresholds and the count rates associated with them.

[0029] Part of the recalibration device according to the invention is also a comparison unit for pixel-by-pixel determination of pixel-specific shift values ​​for the signal threshold values ​​by a comparison between the updated signal threshold value-count value pairs and the pixel-specific reference signal threshold value-count value pairs determined at an earlier time from a previous calibration of the radiation detector.

[0030] The recalibration device according to the invention also has a recalibration unit for pixel-by-pixel determining recalibrated signal threshold-signal energy value pairs by adding the respective pixel-by-pixel shift value to signal thresholds of respective pixel-by-pixel reference signal threshold-signal energy value pairs of the respective pixel from the previous calibration, and for recalibrating the radiation detector using the recalibrated signal threshold-signal energy value pairs. The recalibration device according to the invention shares the advantages of the inventive method for recalibrating a radiation detector with a plurality of pixels.

[0031] The radiation detector according to the invention, preferably an X-ray detector, has a detection unit that generates a detection signal for radiation, in particular X-rays, impinging on the detection unit. The radiation detector according to the invention also has an evaluation unit that, based on the detection signal, determines a set of measured values ​​for X-rays impinging on the detection unit. The measured values ​​can, for example, relate to counting rates of counting events recorded with the aid of the radiation detector. They can also relate to time intervals between individual counting events, counting values, intensity values, or photon rates in a detector pixel. The set of measured values ​​usually comprises, as mentioned above, several counting rates that are determined simultaneously for a single detector channel.The count rates determined simultaneously for a single detector channel are determined, in particular for multiple energy thresholds, with energy threshold values ​​assigned to each of them that differ from one another. Furthermore, the radiation detector according to the invention also comprises a recalibration device according to the invention.

[0032] Such a radiation detector can be designed, for example, as a quantum-counting or photon-counting X-ray detector. In this case, the aforementioned detection unit of the X-ray detector according to the invention can, in particular, comprise a detection region for generating, for example, charge pulses as a detection signal for X-rays absorbed by the detection unit. The detection signal and, in particular, the charge pulses are thus particularly suitable for determining the energy of the absorbed X-rays.

[0033] The evaluation unit can, in particular, be integrated into a so-called ASIC (Application Specific Integrated Circuit) or implemented in the form of an ASIC. In particular, the detection unit and at least parts of the evaluation unit can be included in the aforementioned detector channel. The radiation detector according to the invention shares the advantages of the recalibration device according to the invention.

[0034] The invention further proposes an imaging system, preferably an X-ray imaging system, in particular a CT system, comprising a radiation source, preferably an X-ray source for emitting X-rays, and a radiation detector according to the invention, preferably an X-ray detector, for detecting the radiation emitted by the radiation source. The imaging system according to the invention shares the advantages of the radiation detector according to the invention.

[0035] A large portion of the aforementioned components of the recalibration device according to the invention can be implemented entirely or partially in the form of software modules in a processor of a corresponding computing system, for example, a control unit of a radiation detector or an imaging system with such a radiation detector. This particularly applies to the comparison unit and the recalibration unit. A largely software-based implementation has the advantage that even previously used computing systems can be easily upgraded to operate in the manner according to the invention via a software update.In this respect, the object is also achieved by a corresponding computer program product with a computer program that can be loaded directly into a computing system, with program sections for executing the steps of the method according to the invention, at least the computer-executable steps for determining pixel-by-pixel displacement values, for determining recalibrated signal threshold-signal energy value pairs, and for recalibrating the radiation detector using the recalibrated signal threshold pairs, when the program is executed in the computing system. Such a computer program product can, in addition to the computer program, optionally include additional components such as documentation and / or additional components, including hardware components such as hardware keys (dongles, etc.) for using the software.

[0036] In particular, the recalibration device according to the invention can be part of a user terminal or a control device of an imaging system, preferably an X-ray imaging system, in particular a CT system.

[0037] A computer-readable medium, such as a memory stick, a hard disk, or another portable or permanently installed data storage device, can be used for transport to the computer system or control device and / or for storage on or in the computer system or control device. The computer system can store the program sections of the computer program that can be read and executed by a computer system. For this purpose, the computer system can, for example, have one or more cooperating microprocessors or the like.

[0038] Further, particularly advantageous embodiments and developments of the invention emerge from the dependent claims and the following description, wherein the independent claims of one claim category can also be developed analogously to the dependent claims of another claim category.

[0039] In one embodiment of the method according to the invention, the radiation detector comprises one of the following detector types: - an X-ray detector, - a quantum-counting X-ray detector, - a photodetector.

[0040] The method according to the invention can advantageously be applied to different types of radiation detectors. However, it is particularly preferred to apply the method according to the invention to a quantum-counting X-ray detector.

[0041] In a preferred embodiment of the method according to the invention for recalibrating a radiation detector with a plurality of pixels, the pixel-by-pixel determination of the pixel-individual displacement values ​​is carried out on the basis of one of the following comparison method types: - a regression analysis is carried out on the basis of the updated signal threshold-count pairs and the pixel-specific reference signal threshold-count pairs determined at an earlier time from a previous calibration of the radiation detector, - a method based on artificial intelligence is applied to the updated signal threshold-count pairs and the pixel-specific reference signal threshold-count pairs determined at an earlier time from a previous calibration of the radiation detector.

[0042] Statistical errors, especially noise effects, can be advantageously reduced by using the entire signal threshold scale or DAC value scale to determine the shift.

[0043] In one embodiment of the method according to the invention for recalibrating a radiation detector with a plurality of pixels, the pixel-specific shift values ​​each comprise an offset value. The offset value is added to the signal thresholds of the respective pixel-specific reference signal threshold-signal energy value pairs of the respective pixel of the previous calibration during the step of determining recalibrated signal threshold-signal energy value pairs pixel by pixel. Such a recalibration corresponds to a zero-order recalibration, which is particularly simple and requires little effort to perform.

[0044] If the recalibration is to be even more precise, it can also be performed at a higher order than the aforementioned zeroth order. A compensation curve is determined between the signal threshold values ​​determined during recalibration and the corresponding reference signal threshold values ​​from an earlier calibration. The offset values ​​are determined based on the determined parameter values ​​of the compensation curve and may depend on the respective signal threshold values. The compensation curve preferably includes a polynomial of a higher order than the zeroth order.

[0045] For example, in a first-order recalibration, the updated recalibrated pixel-related shift values ​​are calculated from an offset value and a change in the slope of a best-fit line, preferably determined in a regression analysis. The best-fit line represents the relationship between the signal thresholds measured during recalibration and the signal thresholds of the reference measurement or older calibration.

[0046] The change in the slope of the best-fit line results from the change in its slope compared to a reference value, preferably the value 1. A shift value results from the sum of the offset value and a product of the change in the slope of the determined best-fit line and a signal threshold of a reference-signal threshold-count value pair. Advantageously, the accuracy of the recalibration is further improved by taking the change in the slope into account compared to a zero-order recalibration, in which only the offset value is considered.

[0047] The determined shift can be applied to a plurality of different calibration tables based on signal thresholds during the recalibration step using the recalibrated signal threshold-signal energy value pairs. This is particularly possible because, unlike a complete recalibration, no new noise effects are generated during recalibration. As explained in detail below, in addition to the energy adjustment table, which represents a mapping between signal thresholds and signal energy values ​​or provides correction values ​​for this mapping, there are also other types of calibration tables based on signal thresholds that provide corrections for measured values ​​from radiation detectors.Advantageously, the effort required for recalibration is particularly reduced when recalibrating a plurality of calibration tables of different types compared to a complete recalibration of these calibration tables.

[0048] The different calibration tables preferably include at least one of the following types: - an energy adjustment table, - a linearity correction table, - an offset correction table, - a gain correction table, - a defect pixel correction table, - a hardening correction table.

[0049] An energy adjustment table includes the relationship between DAC values, i.e. signal threshold values, and signal energy values ​​already described in the introduction.

[0050] A linearity correction table is used to correct so-called pileup effects and general nonlinear effects of the measurement scale. Pileup effects involve the measurement of events that occur within such a short period of time that they are hardly perceived as individual events. This results in a counting loss.

[0051] An offset correction table enables zero point correction of pixel-individual differences in the measurement of X-rays.

[0052] A gain correction table enables scaling correction of pixel-specific differences in the measurement of X-rays.

[0053] A defect pixel correction table enables the detection of insufficiently functional pixels for which replacement interpolation values ​​are generated.

[0054] A hardening correction table allows spectral pixel-specific correction for physical properties in the measurement object.

[0055] Advantageously, the shift values ​​can be applied to the different calibration tables. Thus, the individual tables do not need to be recalibrated or retuned individually through a separate calibration measurement. Thus, with a correspondingly large number of different calibration tables, the method according to the invention achieves considerable time savings and a significant reduction in measurement and computational effort.

[0056] In a preferred embodiment of the method according to the invention for recalibrating a radiation detector with a plurality of pixels, the pixel-individual reference signal threshold count pairs are assigned to a continuous reference signal threshold scale and have continuous reference signal thresholds.

[0057] To determine the recalibrated signal thresholds, the determined pixel-individual shift value is added to the continuous reference signal thresholds, whereby continuous signal thresholds are generated, and recalibrated signal thresholds of the recalibrated signal threshold-count pairs are determined by rounding the continuous signal thresholds to an integer signal threshold or DAC value.

[0058] Therefore, a correction is preferably only performed for pixels whose previously calibrated signal threshold is close to a value that is an odd multiple of half a predefined step size of the signal threshold scale. For example, this step size can be 1. A correction is then only performed if a recalibrated rounded signal threshold falls below or exceeds the rounding limit of 0.5, 1.5, 2.5, etc., i.e., the signal threshold to be rounded is close to an odd multiple of half the step size, or the recalibration results in rounding to a different multiple of this value than the previous reference signal threshold.

[0059] Thus, the shift is initially applied on a continuous scale, and during adjustment, it is rounded to a "DAC scale" with integer DAC values. However, the method according to the invention also works on a continuous scale.

[0060] If the method according to the invention is executed multiple times at different times and the continuous signal threshold values ​​are saved during each recalibration, a trend analysis is preferably subsequently performed based on the continuous signal threshold values ​​determined at different times. On this basis, an approximate determination of updated recalibrated signal threshold-signal energy value pairs can be carried out even without repeating the method according to the invention. Advantageously, the effort required for recalibration is further reduced.

[0061] Also preferably, the trend analysis is used to determine the point in time at which the inventive recalibration method needs to be performed again. For example, if the displacement exceeds a predetermined threshold, a new recalibration should be performed using the inventive method. The trend analysis allows such a point in time to be estimated in advance, thus limiting the number of recalibrations to a minimum, thus further reducing the calibration effort.

[0062] To obtain particularly reliable calibration results, a validation of the recalibration according to the invention is preferably carried out based on at least one error criterion, preferably based on a plurality of such error criteria. The error criterion or criteria is / are determined based on the measurement performed for recalibration.

[0063] Based on the validation, it can be determined that if an excessive shift occurs, no recalibration or only a partial recalibration of selected pixels should be performed. An excessive shift should, if necessary, be considered a reason to perform a complete retuning. Recalibration can advantageously be applied specifically to cases where it represents an improvement over the previous calibration.

[0064] Preferably, the at least one error criterion comprises at least one of the following types of criteria: - a shift value occurs which exceeds a predetermined threshold, - a number of pixels for which recalibrated signal threshold-signal energy value pairs must be generated exceeds a predetermined maximum proportion of the total number of pixels of the radiation detector, - a number of pixels classified as defective exceeds a predetermined maximum number.

[0065] The predetermined threshold of the shift depends on the granularity of the signal threshold scale used. For example, with a granularity of 0.5 keV, the predetermined threshold of the shift is preferably two DAC values.

[0066] The above review is carried out in an alternative form using an AI-based process (AI stands for "artificial intelligence"). Advantageously, different criteria can be taken into account in a particularly customized manner.

[0067] If the validation result is negative, a warning message is preferably issued. This is advantageous for informing the user that recalibration has not been performed and warning them of a potential deviation and inaccuracy in a measurement.

[0068] In a particularly advantageous variant of the method according to the invention, the different electronic thresholds of the same detector pixel are mapped onto a common DAC scale, i.e., electronic signal threshold scale. In X-ray detectors, multiple electronic signal thresholds can be assigned to a single pixel. These can be used to cover the entire bandwidth of the X-ray energy spectrum. If a common DAC scale is desired for all signal thresholds, the individual signal threshold scales must be coordinated with one another. This can be achieved, for example, by shifting, compressing, or stretching the individual scales. Advantageously, with such a common DAC scale, all measured count rates of a detector pixel can be represented in a graph. Advantageously, the common DAC scale can be correctly related to an energy scale using the method according to the invention.

[0069] In a particularly effective variant of the method according to the invention, the detector pixels are divided into a first and a second group. The first group of detector pixels is selected for recalibrating the radiation detector. The detector pixels included in the first group are selected according to a certain sorting criterion, which selects detector pixels particularly suitable for recalibrating the radiation detector. The second group then forms the remaining detector pixels. Calibration is advantageously carried out based on the suitable detector pixels, thus achieving a more precise result. The sorting criterion used can, for example, consist of excluding detector pixels located close to a scattered radiation grid, a sensor edge, or a supply cable.The detector pixels assigned to the first group are also referred to below as "clean" pixels, which do not experience any shadowing because, in this special case, they are not located close to the scatter grid. Using such a subset of the available pixels enables the determination of unbiased measured values, such as counts, which, using the parameters obtained beforehand, enable the calibration of a signal threshold scale with an energy scale. This solves the frequently occurring problem of a poorly aligned focal point or the influence of possibly suboptimally aligned collimators.

[0070] Particularly preferred for recalibration are ensembles of detector pixels with a sufficiently uniform response behavior of the individual detector pixels. The ensembles of detector pixels are preferably formed based on several neighboring detector pixels with a sufficiently uniform response behavior of the individual detector pixels. In this case, the input variables used for recalibration are not the measured values ​​of individual detector pixels, but rather values ​​based on a statistical evaluation of the ensembles.

[0071] For example, an ensemble of detector pixels is assigned an ensemble average of the measured values ​​of the individual detector pixels in the pixel group. The measurement signals of these pixels can be summed, for example, to obtain improved statistical significance compared to the measurement of a single pixel. Grouping the pixels into pixel groups significantly increases the signal-to-noise ratio of the measured values ​​obtained during the test measurement and therefore improves the quality and reliability of the recalibration. Furthermore, grouping pixels into pixel groups also reduces the computational effort for recalibration, as the total number of sensor subunits to be recalibrated is reduced compared to recalibrating individual pixels.

[0072] Even within the pixels of such a pixel group, the individual threshold values ​​of the individual detector pixels can be aligned to a common scale. Such alignment is less sensitive to the limited statistics of the individual measured values ​​of the underlying test measurement and thus enables the aggregation of the measured values ​​of a pixel group in the first place.

[0073] To also consider the detector pixels that do not belong to the first group of detector pixels, i.e., those that do not exhibit approximately the same response behavior, an alignment step is preferably performed. This step is also known in the literature as a homogenization process. During the alignment step, the response behavior of the first and second groups of detector pixels is aligned with the response behavior of the first group of detector pixels. Methods for aligning the detector pixels are described, for example, in DE 10 2006 022 596 A1 and DE 10 2011 080 656 B4, which are hereby incorporated by reference.

[0074] In connection with the method according to the invention, a compensation metric generated during homogenization can be used to determine the shift in the association between DAC values ​​or signal threshold values ​​and signal energy values, since neighboring pixels presumably have approximately a similar shift in the first order and an additional pixel-individual shift component in the higher order.

[0075] In the method described in DE 10 2006 022 596 A1 for adjusting the response behavior of detector pixels, a first measurement is carried out using radiation having a first energy spectrum, wherein a first energy threshold lying within the first energy spectrum exists for each detector pixel. The first count rates determined in this way for X-rays below or above the first energy threshold are recorded for each of the detector pixels. X-rays with a second energy spectrum are then emitted, wherein the first energy threshold remains unchanged. In turn, second count rates for X-rays below or above the first energy threshold are recorded for each of the detector pixels. A quotient of the first count rate and the second count rate is then calculated for each of the detector pixels.The initial energy thresholds of the detector pixels are then adjusted to equalize the quotients for the individual detector pixels. Advantageously, the recalibration measurement of the method according to the invention can now be used simultaneously to equalize the response behavior of the detector pixels. Thus, the method according to the invention can save effort and time due to the dual use of the measurement data from the recalibration measurement.

[0076] In order to be able to carry out the described alignment step of the detector pixels, the method according to the invention preferably comprises a measurement additional to the recalibration measurement, which is carried out with a photon flux spectrum different from the recalibration measurement. This second measurement then corresponds to the described measurement with a second energy spectrum within the scope of the alignment method described in DE 10 2006 022 596 A1.

[0077] The X-ray detector according to the invention can in particular comprise a detection region or a detection surface which has a semiconductor material which directly detects or absorbs X-rays (direct converter).

[0078] It is also conceivable for the X-ray detector according to the invention to comprise a detection region with scintillator material that converts X-rays into radiation in other, particularly visible, spectral ranges. The converted radiation can be detected by a semiconductor detector, such as a photodiode or a silicon photomultiplier, arranged downstream of the scintillator in the radiation path and usually also included in the detection unit. The photodiode, the silicon photomultiplier, or the semiconductor detector then generates a detection signal, which can also be analyzed with regard to the spectral distribution and intensity of the X-rays detected by the scintillator.

[0079] The invention will be explained in more detail below with reference to exemplary embodiments in the accompanying figures. In the various figures, identical components are provided with identical reference numerals. They show: Fig. 1 a diagram illustrating the drift of an X-ray detector using the example of measuring a characteristic X-ray spectrum of an X-ray source, Fig. 2 is a diagram illustrating a first and a second measurement curve of signal threshold count rate pairs recorded at different times, Fig. 3 is a graph illustrating updated signal thresholds plotted against reference signal thresholds, Fig. 4 is a graph showing distributions of pixels with shifts for four electronic energy thresholds for the zeroth order of the shifts, Fig. 5 is a graph showing distributions of pixels with shifts for four electronic energy thresholds for the first order of the shifts, Fig. 6 is a flowchart illustrating a method for recalibrating a radiation detector having a plurality of pixels according to a first embodiment of the invention, Fig. 7 is a flowchart showing step 6.IV of the Fig. 6 according to a first embodiment of the invention, Fig. 8 is a block diagram illustrating a calibration device according to an embodiment of the invention, Fig. 9 is a block diagram illustrating an X-ray detector with a calibration device according to an embodiment of the invention.

[0080] In Fig. Figure 1 shows a diagram 10 illustrating the drift of an X-ray detector using the example of measuring a characteristic X-ray spectrum of an X-ray source. A first measurement curve m1, which shows a normalized spectral response nsr plotted against signal energy values ​​E[keV], illustrates the measurement properties of the X-ray detector at a first time, and a second measurement curve m2, which is slightly shifted from the first measurement curve m1, illustrates the measurement properties of the X-ray detector at a second time, which is later than the first time. Fig. 1 also shows a number of prominent energy values ​​of the X-ray spectrum, such as 23 keV (fluorescence of the X-ray detector), 59 keV (fluorescence of the X-ray tube) and 68 keV, which are shown in the diagram of the Fig. 1 maxima are assigned. These maxima or the energy values ​​assigned to them can be used as orientation values ​​in a conventional calibration. However, noise effects often make it difficult to determine the maxima precisely. Due to the aforementioned noise effects, a certain number of pixels are conventionally combined for spectral fine adjustment, since only by combining and averaging over a plurality of pixels can the Fig. 1 illustrated spectral structures become visible.

[0081] In Fig. Figure 2 shows a graph 11 illustrating a first measurement curve M1 and a second measurement curve M2 of signal threshold-count rate pairs recorded at different times. It is clearly visible that, for the same count rate value ZR, a shift in the associated signal threshold values ​​DAC occurs.

[0082] In Fig. 3 shows a graph 12 which shows updated signal thresholds DAC over reference signal thresholds DAC ref For this purpose, the Fig. 2 curves are figuratively superimposed and DAC values ​​or signal threshold values ​​DAC assigned to the same counting rates are ref , DAC are compared with each other. The angle bisector symbolizes a shift-free mapping of the reference signal threshold values ​​DAC ref to the updated DAC energy thresholds. As in Fig. 3, the new signal threshold values ​​DAC differ slightly from the reference signal threshold values ​​DAC ref This deviation has an average value of V = -0.148 DAC in the zero order. This means that the shift during recalibration is -0.148 DAC values.

[0083] In Fig. Figure 4 shows a graph 13 illustrating distributions ZP of pixels P with displacements V and ΔDAC for four electronic energy thresholds th1, ..., th4 for the zeroth order of the displacements V. The displacements are characterized by offset values ​​ranging between -2 and +1.

[0084] In Fig. 5 shows a diagram 14 which depicts distributions ZP of pixels with shifts V or ΔDAC of the first to fourth energy thresholds th1, th2, th3, th4 for the first order of the shifts with two sub-diagrams. In the left sub-diagram, offset values ​​of the shift V can be seen. The shifts are characterized by offset values ​​that lie between -5 and +4. In the right-hand diagram in Fig. 5 is a deviation ΔS of the slope of a regression line between the reference energy thresholds DAC ref and the newly measured energy threshold values ​​DAC.

[0085] In Fig. 6 shows a flowchart 600 illustrating a method for recalibrating a radiation detector 1 having a plurality of pixels P according to a first embodiment of the invention.

[0086] In step 6.I, a characteristic X-ray spectrum of an X-ray source is detected by the radiation detector 1. For each pixel P of the radiation detector 1, a plurality of energy-dependent counting rates ZR are measured for different signal threshold values ​​DAC of an evaluation unit of the radiation detector 1.

[0087] In step 6.II, updated signal threshold-count pairs SZP are determined by assigning the signal thresholds DAC to spectral count rates ZR based on the knowledge of the characteristic X-ray spectrum.

[0088] In step 6.III, pixel-by-pixel individual displacement values ​​V ifor the signal threshold values ​​DAC by a comparison between the signal threshold count pairs SZP with pixel-individual reference signal threshold count pairs SZP determined at an earlier time ref a previous calibration of the radiation detector 1 based on a regression analysis.

[0089] In step 6.IV, recalibrated signal threshold-signal energy value pairs SEP neu by pixel-by-pixel adding of the respective pixel-individual displacement value V i to the signal thresholds DAC ref the respective pixel-individual reference signal threshold-signal energy value pairs SEP ref of the respective pixel P i Furthermore, the radiation detector 1 is recalibrated using the recalibrated signal threshold-signal energy value pairs SEP neu .

[0090] In Fig. Figure 7 shows a flowchart illustrating step 6.IV according to a first embodiment of the invention.

[0091] In step 6.IVa, pixel-individual reference signal threshold-signal energy value pairs SEP ref assigned to a continuous reference signal threshold scale so that they provide continuous reference signal thresholds DAC ref - _kont have.

[0092] In step 6.IVb, the determined pixel-individual displacement value V i to the continuous reference signal thresholds DAC Ref_kont added, with continuous signal thresholds DAC neu_kont be generated.

[0093] Subsequently, in step 6.IVc, recalibrated signal threshold values ​​DAC neu by rounding the continuous signal threshold values ​​DAC neu kont to an integer signal threshold DAC ganzThese integer recalibrated signal threshold values ​​DAC neu are each part of a recalibrated signal threshold-signal energy value pair SEP neu .

[0094] In Fig. 8 is a recalibration device 50 for a radiation detector 1 (see Fig. 9) according to an embodiment of the invention. The recalibration device 50 comprises a data reception interface 51 for receiving a characteristic X-ray spectrum of an X-ray source. For each pixel P of the radiation detector 1, a plurality of energy-dependent counting rates ZR are measured for different signal threshold values ​​DAC of an evaluation unit of the radiation detector 1.

[0095] Part of the recalibration device 50 is an assignment unit 52 which is configured to determine updated signal threshold-count pairs SZP by assigning the signal threshold values ​​DAC to spectral count rates ZR on the basis of the characteristic X-ray spectrum.

[0096] The recalibration device 50 also comprises a comparison unit 53 for pixel-by-pixel determination of pixel-individual displacement values ​​V i for the signal threshold values ​​DAC by a comparison between the signal threshold count pairs EZP with pixel-individual reference signal threshold count pairs SZP determined at an earlier time ref a previous calibration of the radiation detector 1 based on a regression analysis.

[0097] The recalibration device 50 according to the invention also has a recalibration unit 54 for pixel-by-pixel determination of recalibrated signal threshold value-signal energy value pairs SEP neu by pixel-by-pixel adding of the respective pixel-individual displacement value V i to the signal thresholds DAC ref the respective pixel-individual reference signal threshold count pairs SZP zef of the respective pixel P i on.

[0098] The determined recalibrated signal threshold-signal energy value pairs SEP neu are finally transmitted to an output interface 55, which sends these recalibration data SEP neu to a database DB (see Fig. 9) in which one or more calibration tables are stored.

[0099] In Fig. Figure 9 schematically shows a radiation detector 1 in the form of a semiconductor detector that functions as a direct converter. Such a semiconductor detector 1 is used, for example, in a computed tomography scanner. The semiconductor detector 1 is designed as a planar pixel detector. It comprises semiconductor material 2 in the form of a single crystal as the detector material. The semiconductor material 2 is covered on one side with a metal surface that forms an electrical rear contact 3. A voltage HV is applied to the rear contact 3.

[0100] The opposite side of the single crystal 2 is covered with a structured metallization, the individual partial surfaces of which are designed as pixel contacts 4 and, as a whole, form a pixel matrix. The size and spacing of the pixel contacts 4, together with other material parameters of the semiconductor material 2, determine the maximum spatial resolution of the semiconductor detector 1. The resolution is typically in the range of 10 to 500 µm. The individual pixel contacts 4 are each connected to separate readout electronic circuits 5, which detect measurement signals s1, s2, s3, s4. The comparators built into the readout electronic circuits 5 can be adjusted to detect specific spectral components of X-rays using electronic threshold voltages.

[0101] Furthermore, the radiation detector 1 comprises an addition element 6, with which the measurement signals s1, s2, s3, s4 are combined into a group signal G. The addition element 6 can also perform a weighting of the individual measurement signals or pixel signals before adding them. In order to correctly add the individual signals s1, s2, s3, s4, the electronic thresholds of the comparators assigned to them must be set such that they correspond exactly to the same energy threshold. For this purpose, the Fig. 8 is used. The recalibration device 50 sends out the signal as part of the process associated with Fig. 6 and Fig. 7, threshold setting commands SEB are sent to the comparators 5 in order to change or ultimately set their comparator threshold and receive measurement signals ZR from the individual comparators 5 (only the signal connection to the comparator transmitting a fourth signal s4 is shown; however, analogous signal connections are also present to the other comparators). Furthermore, the recalibration device 50 can optionally also control an X-ray tube R as part of the recalibration with the aid of control commands SB, whereby the X-ray flux for the individual measurements is changed accordingly. This procedure can be useful if an adjustment or homogenization of pixels is to take place. After the calibration has been completed, the determined calibration data KD are transmitted to a database DB.The calibration data KD can be used, for example, when processing the counting rates ZR of the individual pixels in the addition element 6 in order to correctly add the counting rates ZR according to the correctly assigned energy values.

[0102] It is noted that arrangement 1 in Fig. 9 is shown as a cross section and therefore the arrangement 1 in Fig. 9 comprises a total of 16 pixel contacts. For clarity, only 4 pixel contacts per pixel group were drawn, corresponding to the cross-sectional view. As already mentioned, 16 or a different number of pixel contacts per pixel group is common.

[0103] The Fig.The semiconductor detector 1 shown in Figure 9 functions as follows: X-rays incident on the semiconductor detector 1 cause interactions of the X-ray quanta with the semiconductor material 2 of the semiconductor detector 1, generating electron-hole pairs. The voltage HV applied to the contacts of the detector 1 generates an electric field, which moves the generated charge carriers towards the electrical contacts, in particular towards the pixel contacts 4. The charge movement in the sensor material between the electrodes or pixel contacts 4 of the semiconductor detector 1 generates an electric charge pulse that is proportional to the absorbed energy of the X-ray radiation. This charge pulse is read out by the connected readout electronic circuit 5. The pixel signals or measurement signals s1, s2, s3, s4 detected by the readout electronic circuit 5 are forwarded to an addition element 6, which adds the measurement signals s1, s2, s3, s4 (actually the measurement signals s1 to s16 ) to form a group signal G. The number of four different measurement signals s1, s2, s3, s4 is merely exemplary and can be replaced by any number of n measurement signals.

[0104] Finally, it should be noted that the features of all exemplary embodiments or developments disclosed in the figures can be used in any combination. Finally, it should also be noted that the X-ray detector, the X-ray imaging system, the calibration device, and the method for recalibrating a radiation detector described in detail above are merely exemplary embodiments that can be modified in a variety of ways by a person skilled in the art without departing from the scope of the invention. For example, the X-ray detector can also be designed as an annular detector that completely encloses a measuring space of an imaging system in one direction. Furthermore, the use of the indefinite articles "a" or "an" does not exclude the possibility that the features in question can also be present multiple times.Likewise, the term "unit" does not exclude the possibility that the components in question consist of several interacting subcomponents, which may also be spatially distributed. Regardless of the grammatical gender of a particular term, it includes persons with male, female, or other gender identities.

Claims

[1] Method for recalibrating a radiation detector (1) having a plurality of pixels (P i ), comprising the steps: - detecting a radiation spectrum of a radiation source by the radiation detector (1), wherein for each pixel (P i ) of the radiation detector (1), a plurality of energy-dependent counting rates (ZR) for different signal threshold values ​​(DAC) of an evaluation unit (5) of the radiation detector (1) are measured and updated signal threshold value-counting value pairs (SZP) are determined on the basis of the different signal threshold values ​​(DAC) and the counting rates (ZR) assigned to them, - pixel-by-pixel determination of pixel-individual displacement values ​​(V i ) for the signal thresholds (DAC) by comparing the updated signal threshold count pairs (SZP) with pixel-specific reference signal threshold count pairs (SZP ref) a previous calibration of the radiation detector (1), - pixel-by-pixel determination of recalibrated signal threshold-signal energy value pairs (SEP neu ) by adding pixel by pixel the respective pixel-specific displacement value (V i ) to signal thresholds (DAC ref ) of respective pixel-individual reference signal threshold-signal energy value pairs (SEP ref ) of the respective pixel (P i ) of the previous calibration, - Recalibrating the radiation detector (1) using the recalibrated signal threshold-signal energy value pairs (SEP neu ) [2] Method according to claim 1, wherein the pixel-by-pixel determination of the pixel-individual displacement values ​​(V i ) is based on one of the following types of settlement procedure: - a regression analysis is carried out on the basis of the updated signal threshold count pairs (SZP) and the pixel-specific reference signal threshold count pairs (SZP ref ) a previous calibration of the radiation detector (1) is carried out or - a method based on artificial intelligence is applied to the updated signal threshold count pairs (SZP) and the pixel-specific reference signal threshold count pairs (SZP ref ) a previous calibration of the radiation detector (1). [3] Method according to claim 1 or 2, wherein the pixel-individual displacement values ​​(V i ) each comprise an offset value (ΔDAC) which, in the step of pixel-by-pixel determining of recalibrated signal threshold-signal energy value pairs (SEP neu ) to signal thresholds (DAC ref) of respective pixel-individual reference signal threshold-signal energy value pairs (SEP ref ) of the respective pixel (P i ) is added to the previous calibration. [4] Method according to one of the preceding claims, wherein the pixel-related displacement values ​​(V i ) each based on a determined function of higher order than the zeroth order between the signal thresholds (DAC) of the updated signal threshold-count pairs (SZP) and the signal thresholds (DAC ref ) of the reference signal threshold count pairs (SZP ref ) can be determined. [5] Method according to one of the preceding claims, wherein a plurality of different calibration tables based on signal threshold values ​​(DAC) are used in the step of recalibrating by means of the recalibrated signal threshold-signal energy value pairs (SEP neu ) based on the determined displacement values ​​(Vi ) be corrected. [6] Method according to one of the preceding claims, wherein - the pixel-individual reference signal threshold count pairs (SZP Ref ) are assigned to a continuous reference signal threshold scale and continuous reference signal thresholds (DAC Ref_kont ), - first, the determined pixel-individual displacement value (V i ) to the continuous reference signal thresholds (DAC Ref _ kont ) is added, whereby continuous signal thresholds (DAC neu_kont ) are generated, and - recalibrated signal thresholds (DAC neu ) of the recalibrated signal threshold-signal energy value pairs (SEP neu ) by rounding the continuous signal thresholds (DAC neu_kont ) to an integer signal threshold value (DAC ganz ) can be determined. [7] Method according to claim 6, wherein the method is carried out several times at different times and during each recalibration the continuous signal threshold values ​​(DAC neu_kont ) and subsequently calculated based on the continuous signal threshold values ​​(DAC neu_kont ) a trend analysis is carried out, - on the basis of which an approximate determination of updated recalibrated signal threshold-signal energy value pairs (SEP neu ) can also be carried out without repeating the process according to one of claims 1 to 6, and / or - on the basis of which it is determined at what point in time the method according to one of claims 1 to 6 must be carried out again. [8] Method according to one of the preceding claims, wherein a validation of the recalibration is carried out on the basis of at least one error criterion based on the measurement carried out for the recalibration, preferably on the basis of a plurality of such error criteria. [9] The method of claim 8, wherein the at least one error criterion comprises at least one of the following types of criteria: - a shift value (V, V i ) which exceeds a predetermined threshold, - a number (Z i ) of pixels (P i ), for which recalibrated signal threshold-signal energy value pairs (SEP neu ) must be generated, exceeds a predetermined proportion (A max ) the total number of pixels (P) of the radiation detector (1), - a number (Z k ) of pixels classified as defective (P k ) exceeds a predetermined number (Z max ). [10] Method according to claim 8 or 9, wherein a warning message is issued in the event that the result of the validation is negative. [11] Recalibration device (50) for a radiation detector (1), comprising: - a data reception interface (51) for detecting a characteristic radiation spectrum of a radiation source by the radiation detector (1), wherein for each pixel (P i ) of the radiation detector (1), a plurality of energy-dependent counting rates (ZR) for different signal threshold values ​​(DAC) of an evaluation unit (5) of the radiation detector (1) are received and updated signal threshold value-counting value pairs (SZP) are determined on the basis of the different signal threshold values ​​(DAC) and the counting rates (ZR) assigned to them, - a comparison unit (53) for pixel-by-pixel determination of pixel-individual displacement values ​​(V i) for the signal thresholds (DAC) by comparing the updated signal threshold count pairs (SZP) with the pixel-specific reference signal threshold count pairs (SZP ref ) a previous calibration of the radiation detector (1), - a recalibration unit (54) for pixel-by-pixel determination of recalibrated signal threshold-signal energy value pairs (SEP neu ) by adding pixel by pixel the respective pixel-specific displacement value (V i ) to the signal thresholds (DAC Ref ) of respective pixel-individual reference signal threshold count pairs (SEP Ref ) of the respective pixel (P i ) of the previous calibration and for recalibrating the radiation detector (1) using the recalibrated signal threshold-signal energy value pairs (SEP neu ) [12] Radiation detector (1), preferably X-ray detector, comprising: - a detection unit (2) which generates a detection signal for radiation, preferably X-rays, incident on the detection unit (2), - an evaluation unit (5) which determines a plurality of energy-dependent counting rates (ZR) for different signal threshold values ​​(DAC) on the basis of the detection signal, - a recalibration device (50) according to claim 11 for recalibrating the radiation detector (1). [13] Imaging system, preferably X-ray imaging system, in particular CT system, with a radiation source (R) and a radiation detector (1) according to claim 12. [14] Computer program product, comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps for determining pixel-by-pixel displacement values ​​(V i), for pixel-by-pixel determination of recalibrated signal threshold-signal energy value pairs (SEP neu ) and for recalibrating the radiation detector (1) by means of the recalibrated signal threshold value-signal energy value pairs (SEP neu ) of the method according to one of claims 1 to 10. [15] Computer-readable storage medium, comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps for determining pixel-by-pixel displacement values ​​(V i ), for pixel-by-pixel determination of recalibrated signal threshold-signal energy value pairs (SEP neu ) and for recalibrating the radiation detector (1) by means of the recalibrated signal threshold value-signal energy value pairs (SEP neu ) of the method according to one of claims 1 to 10.

Images