X-ray detection system

Abstract

The invention relates to an X-ray detection system comprising a detector unit (3) for generating an input signal (IS) on the basis of detected X-ray photons, and a signal processing device (11) for processing the input signal, in which the signal processing device (11) comprises:-input means for receiving the input signal (IS), -processing means for generating an output signal (OS) from the input signal (IS),-control means for influencing the input signal (IS),-the control means being capable of determining a statistical value of the output signal (OS), comparing the statistical value with a predetermined setpoint value, and influencing the input signal, such that a difference between the statistical value and the setpoint value in the output signal is reduced.

Description

Technical Field

[0001] This invention relates to an X-ray detection system including a signal processing device, the signal processing device comprising: an input device for receiving an input signal based on detected X-ray photons; a processing device for generating an output signal from the input signal; and a control device for influencing the input signal based on a determined statistical value of the output signal. Furthermore, this invention relates to a photon-counting X-ray imaging system including such an X-ray detection system. Additionally, this invention relates to a computed tomography apparatus and a method for processing signals. Background Technology

[0002] Typically, the task of photon counting systems used in medical X-ray imaging, as well as other particle detectors, is to detect and count pulses generated by recorded particles, and usually also to determine their energies. For this purpose, sensor materials (e.g., CdTe, CdZnTe, Si, ...) are used in direct conversion detectors, where, for example, electron / hole pairs are generated when photons are absorbed. With the aid of a depletion voltage (also known as a bias voltage), the electron / hole pairs are separated and directed to opposing electrodes of the sensor. This affects a current signal on the electrodes, which is typically detected by electronic circuitry on one side of the pixelated electrodes. This amplifies the charge pulses, performs pulse shaping if necessary, and then compares the signal with at least one comparator. Thus, when absorbed in a direct conversion sensor, each X-ray quantum generates a short current pulse whose size / total charge is proportional to the energy deposited.

[0003] When the set comparator threshold is exceeded, photons are finally detected and counted. Simultaneously, the minimum energy a photon must possess to be detected is determined by selecting a threshold. The energy of the detected photon can be determined specifically by using multiple comparators with different thresholds.

[0004] The following problem arises when correctly determining the photon energy:

[0005] Because the sequence of X-ray quanta is random, overlap of individual pulses (so-called "pileup") occurs more frequently, especially at high fluxes. This presents a specific technical problem of making the energy classification of pulse measurements as stable and robust as possible over a wide range of possible X-ray fluxes.

[0006] Applying a depletion voltage to the sensor material leads to leakage or dark current, even without incoming photons. This depends on various parameters (sensor material, electrode type and material, sensor thickness, voltage difference, temperature, current, and previous radiation intensity, etc.) and is typically affected by slow fluctuations. Using a transimpedance amplifier, the low-frequency signal component will offset the amplifier's output signal; if the comparator threshold remains constant, this will cause a shift in the detected photon energy, which must therefore be compensated for. Using a charge-sensitive preamplifier, changes in leakage current lead to suboptimal compensation, thus also causing distortion of the detected photon energy or even amplifier saturation.

[0007] If photon flux is present, the actual signal is superimposed on the dark current. In the case of single pulses with sufficient time intervals, each pulse will start at the baseline (the resting state without pulses) and return completely. Therefore, the pulse height and photon energy can be correctly determined. However, as photon flux increases, low-frequency signal components also increase, making it difficult to compensate for dark current alone.

[0008] Furthermore, pulses overlap increasingly frequently, i.e., they overlap (a so-called stacking situation) and can be identified as photons with higher energies. As a result, the count rate spectrum, representing the count events per time as a function of photon energy, is erroneously shifted relative to the energy spectrum of the X-ray source as a function of photon flux. Therefore, the random sequence of X-ray quanta leads to an increase in the superposition of individual pulses due to stacking, especially at high fluxes. This raises a specific technical problem of making the energy classification of pulse measurements as stable and robust as possible over a wide range of possible X-ray fluxes.

[0009] In particular, in the field of medical X-ray imaging, and even more specifically in computed tomography, when scanning different regions of a target with different absorption characteristics, the photon flux can change rapidly, i.e., from one measurement interval to the next. If the signal baseline does not adapt to the changing photon flux quickly enough, this also leads to erroneous measurement results.

[0010] In many typical implementations, the problem remains unresolved or only partially resolved.

[0011] For example, the sensor leakage current can be determined before the actual X-ray measurement and then compensated for by adjusting the quiescent current at the input node (“quiescent leakage current compensation”).

[0012] Alternatively, the compensation current can be dynamically and slowly adjusted (“dynamic leakage current compensation”) by using, for example, a resistive feedback amplifier with an appropriate transfer function. In this case, it makes sense for the feedback path to begin after the first amplifier, thus ensuring that the output signal remains within its dynamic range. The feedback path typically involves resistive (R) and capacitive (C) paths or more complex impedance networks (e.g., Krummenacher feedback with self-cascaded FETs with zero-pole cancellation (compare example WO2024086081A)). Depending on the configuration of R and C, the amplifier is then called a transimpedance amplifier (R determines the gain, C is small enough to be negligible, “TIA”), a charge-sensitive amplifier (R is almost infinite, C determines the gain, “CSA”), and a charge-sensitive amplifier with continuous reset (R is large, controlling the reset time constant, C determines the gain). The transition between types is continuous, and there are also perceptible hybrid forms (R is small enough to determine the gain, C is large enough to ensure a certain charge collection time). In general, it is thus only the feedback impedance that determines the transfer function of the amplifier, and in particular, the low-frequency portion of the transfer function provides compensation for the leakage current of the input sensor.

[0013] Energy classification is typically designed only for low-flow-rate applications. In this case, all pulses begin from a known rest position, and the maximum height above the rest position is a measure of the pulse energy. Bias due to build-up is then typically corrected downstream using numerical corrections (e.g., linear correction or beam hardening correction).

[0014] One possibility is to use bipolar pulse shaping, in which a downstream undershoot compensates for the positive signal pulse. In this case, the net shift of the rest position is zero, causing the energy level to be briefly disturbed by other pulses, but without systematic shift.

[0015] Typically, a dedicated circuit element, known as a baseline restorer (BLR) or baseline hold (BLH), is implemented in particle detectors for dynamic adjustment of the signal baseline. In photon counting, a topology consisting of a preamplifier with a downstream (and capacitively isolated) pulse shaper is frequently found. In this case, a BLH / BLR is necessary to fully establish the defined output level.

[0016] A baseline restorer (BLR) detects pulseless periods and pulls the signal to the desired baseline during these times. To do this, it typically has a high bandwidth. Especially in applications such as medical X-ray imaging, which sometimes has high photon flux, pulseless phases can be correspondingly sparse and difficult to detect. Therefore, the use of a BLR in applications with prolonged high photon flux is only suitable to a limited extent.

[0017] In contrast to a BLR, a baseline hold (BLH) is characterized by a low bandwidth and filters out low-frequency components from the signal to compensate for them. A typical BLH structure works by first amplifying the difference between the amplified and pulse-shaped signal and the desired baseline, and processing it with a nonlinear function if necessary. The result is then low-pass filtered and converted into a current, which is then fed to the detector input for compensation.

[0018] Without using a nonlinear function, the moving average of the signal is essentially determined and compensated. This behavior corresponds to AC coupling of the signal. As a result, the average value of the signal is also compensated for at higher photon fluxes.

[0019] Applying a nonlinear function to the differential voltage between the signal and the desired baseline yields different behavior. For example, in the case of a unipolar pulse, the signal component below the baseline is weighted much more heavily than the component above the baseline. This can be achieved, for example, by limiting the slew rate to be effective only in one direction. This nonlinear behavior shifts the signal in such a way that the negative peak lies on the baseline. This BLH behavior is also known as DC coupling and essentially only compensates for dark current.

[0020] Therefore, various baseline maintainers differ significantly in their response behavior. Known types are...

[0021] - AC-class coupling, where the integral of the deviation of the output signal from the target value is adjusted to zero. The goal here is to make the average deviation zero, very similar to a bipolar pulse shape. In contrast, a unipolar pulse can also be used with a baseline hold variant, and the response time constant can be selected as needed. For intermediate pulse shapes (proportional undershoot), the feedback effect is proportionally divided between the pulse shape and the baseline hold.

[0022] - Feedback with a certain time constant and a response strength proportional to the deviation (similar to RC behavior). The response strength and time constant can be set here, but a trade-off must be made between the impact of high-throughput pulses and the feedback strength in high-throughput cases. Large pulses vary more than small pulses. Furthermore, the severity of the highly correlated effects of pulses depends on the current intensity of the signal flow / accumulation level. Combined, these two factors compromise the robustness of energy classification.

[0023] - Feedback with a constant current intensity causes the signal to return to the baseline with a constant, signal-independent slope, resulting in a triangular pulse shape. This has the advantage of predictable effects on the signal pulse that are independent of pulse height. However, it also causes a time-varying shift in the energy scale. Furthermore, the time until the signal returns to the rest position depends linearly on the level of the previous deflection and does not have a uniform time constant.

[0024] - Asymmetrical response to overshoot or undershoot of the target voltage, particularly a low response to overshoot during signal pulses (e.g., using a large R), but a strong response to undershoot of the target voltage (e.g., using a diode). The aim here is to ensure a minimum output level, specifically set when the useful signal no longer occurs, while minimizing the impact on the pulse in the case of a signal. A drawback is that the energy scale offset caused by stacking is almost or not compensated for at all. This type is very similar in its operation to a baseline restorer. Summary of the Invention

[0025] The purpose of this invention is to provide stable and robust X-ray detection.

[0026] According to the present invention, the objective is achieved by an X-ray detection system and a method for processing signals as described in the independent claim. Furthermore, preferred developments are defined in the dependent claims. In this patent application, nouns and pronouns relating to persons generally do not specify a particular gender.

[0027] In one aspect, an X-ray detection system is provided, comprising: a detector unit for detecting X-ray photons and for generating an input signal based on the detected X-ray photons; and an advantageous signal processing means for processing the input signal.

[0028] Advantageous signal processing apparatuses include an input device for receiving input signals and a processing device for generating output signals from the input signals. The signal processing device transforms the input signal into an output signal. The input device is capable of receiving the input signal. For example, the input device is implemented by an interface for receiving input signals from an external unit. In the case of the X-ray detection system described above, the external unit is a corresponding detector unit capable of detecting X-ray photons and generating an input signal based on the detected photons. The input signal generated by the detector unit is then received by the input device of the signal processing apparatus. In other applications, the external unit may be, for example, a memory unit or a data network. The input device may include one or more memory units.

[0029] A processing device for generating an output signal from an input signal may include an amplifier for amplifying the input signal. Additionally, the processing device may optionally include a shaper for shaping the amplified input signal to obtain the output signal.

[0030] Furthermore, the signal processing device includes a control device for influencing the input signal. For example, the control device can change the offset or add current to the input signal.

[0031] Furthermore, the control device can determine the statistical value of the output signal. The statistical value may relate to the pulses of the output signal. Additionally, the control device can compare the statistical value with a pre-defined setpoint value. In other words, it compares the actual statistical value with a target value. Furthermore, the control device can influence the input signal to reduce the difference between the statistical value in the output signal and the setpoint value. For example, it can change the signal offset in a way that minimizes the difference between the two values.

[0032] For example, the output signal differs from the input primarily in amplification (and shaping). The output signal can also have completely independent offsets. However, the value of the output offset can be affected by changing the input offset. This allows the control loop to minimize the statistical characteristics of the output by controlling the offset on the input.

[0033] Advantageously, the input signal can be optimized relative to a specific statistical value by simply changing the offset of the input signal.

[0034] The detector unit included in the X-ray detection system may include a sensor, such as CdTe, CdZnTe, Si, etc., used as the sensor material in the case of a direct conversion detector. In response to photons absorbed within the sensor, a current signal is generated at the electrodes of the sensor, representing an input signal to an electronic circuit for signal processing, which is connected to the sensor by signaling. This electronic circuit, implemented, for example, as an ASIC (Application-Specific Integrated Circuit), may include the signal processing device described above; that is, the signal processing device and its means for processing and controlling the input signal received by its input device can be implemented in the form of a circuit within the electronic circuit. Furthermore, the sensor and the electronic circuit for processing the signal from the sensor can be pixelated to allow spatially resolved photon detection—for example, by providing pixelated electrodes on at least one side of the sensor, particularly on the sensor side facing the electronic circuit in terms of signal aspect, and by providing pixelated signal processing electronic circuitry. Therefore, a separate signal processing device for the X-ray detection system can be provided for each pixel of the detector unit. Other implementations are also possible. In addition to the proposed signal processing device, such electronic circuitry as described above may also include analog or digital circuitry to process the received signal. In particular, multiple comparators and associated counters can be provided to enable photon counting, and in particular, by providing multiple comparators, energy-resolved measurements of the detected photons are possible.

[0035] According to an embodiment of the invention, the statistical value is a quantile (or average, median, or other statistical moment) associated with the output signal. The quantile (hereinafter used as a representative of all the above statistical values) can be used to describe the probability that the output signal is below a certain (energy) threshold. For example, the predetermined setpoint value is defined as the 30th quantile. In this case, by influencing the device to offset the output signal (the offset is variable), the statistical value of the output signal is shifted towards the predetermined 30th quantile. Therefore, the quantile of the output signal can be adjusted by changing the offset.

[0036] According to another embodiment, the output signal is a continuous signal, and the quantile refers to the proportion of time the output signal is below a predetermined baseline. In this case, the quantile relates to the relative proportion of time the output signal is below a predetermined threshold called the baseline. The relative time below the baseline is simply referred to as TBB. The output signal can be an analog signal or a digital signal. In either case, the TBB can be calculated relative to a specific baseline.

[0037] In another embodiment, the input signal includes a pulse signal. Specifically, the pulse signal can be considered as a pulse sequence. For example, the pulse signal consists of multiple pulses. The input signal may include additional signal components (e.g., noise, leakage, etc.). The pulses can occur randomly. That is, for example, the input signal, i.e., the pulse signal, is generated by detecting X-ray photons. Therefore, the pulse signal can be used to count X-ray photons.

[0038] In another embodiment, the processing device includes a charge-sensitive amplifier or a transimpedance amplifier for amplifying the input signal. The charge-sensitive amplifier converts charge into a corresponding voltage. Conversely, the transimpedance amplifier converts current into voltage. Both amplifiers can be implemented using one or more operational amplifiers.

[0039] According to another embodiment, the control device can add a compensation current to the input signal, wherein the compensation current is based on the output of a corresponding amplifier. Therefore, a feedback loop can be implemented. Specifically, the input signal is amplified by a charge-sensitive amplifier or a transimpedance amplifier, and a compensation current is generated based on the amplified signal. Finally, the compensation current is added to the input signal, thereby completing the loop. Thus, respective feedback control can be achieved.

[0040] According to another embodiment, the compensation current is generated by a baseline hold circuit (in short: a baseline hold or baseline restorer circuit is based on the output signal of the respective amplifier. As mentioned above, the baseline hold has a relatively small bandwidth and filters out low-frequency components from the output signals of the respective amplifiers in order to compensate for them. Conversely, the baseline restorer circuit detects no-pulse times and pulses the signal to the desired baseline during these times. The baseline restorer circuit typically has a much higher bandwidth than the baseline hold. Therefore, the baseline used to determine the TBB can be influenced by either the baseline hold circuit or the baseline restorer circuit.

[0041] According to another embodiment, the baseline hold includes two adjustable current sources of opposite polarity and independent strength, implemented as part of a control device for determining the statistical characteristics of the output signal. These sources provide a current to the integrator (only) when the output signal is below a preset baseline to generate a compensation current, and (only) when the output signal is above the preset baseline to provide another current of opposite polarity to the integrator to generate a compensation current. Specifically, the baseline hold circuit may include an integrator circuit and two adjustable current sources of opposite polarity connected to the integrator input. The output from the integrator forms the basis of the compensation current, wherein one of the adjustable current sources provides current only when the output signal is below the predetermined baseline, and the other adjustable current source provides current only when the output signal is above the predetermined baseline. This means that the compensation current is generated based on the positive and / or negative currents provided to the integrator by the adjustable current sources. In other words, the positive and negative currents do not represent the compensation current, but are used to generate it.

[0042] Adjustable current sources can be implemented in various ways, such as by a series of selectable switchable partial current sources with different current intensities (e.g., powers of 2) or by a potentiometer.

[0043] An integrator is used to integrate currents from two adjustable current sources, with the integrator's output serving as the basis for the compensation current. Typically, the integrator integrates the current and provides a corresponding voltage. This voltage must be converted into the compensation current. A corresponding V / I converter can be provided. Therefore, the integrator integrates a positive current as long as the output signal of the signal processing device is below the baseline, and integrates a negative current as long as the output signal is above the baseline. Thus, the compensation current reduces the distance from the predetermined baseline.

[0044] In another embodiment, the current intensity of one of the adjustable current sources with opposite polarities is proportional to the proportion of time the output signal is above a predetermined baseline, and the current intensity of the other adjustable current source is proportional to the proportion of time the output signal is below the predetermined baseline. That is, for example, the positive current is proportional to the proportion of time the input signal is above the predetermined baseline, and the negative current is proportional to the proportion of time the input signal is below the predetermined baseline. When the ratio (e.g., TBB) is small, a high positive current and a relatively small negative current are provided. Otherwise, if the ratio is high, a relatively small positive current and a relatively high negative current are provided. Therefore, the absolute values ​​of the two currents have the desired relationship of (1 - TBB) / TBB.

[0045] In another aspect of the invention, a photon-counting X-ray imaging system is provided, comprising the X-ray detection system described above, wherein the input signal includes a corresponding pulse signal, and a counting unit is provided for counting pulses of the output signal that exceed a predetermined energy threshold. Therefore, reliable photon counting can be performed by means of the aforementioned signal processing apparatus.

[0046] Specifically, a counting unit may include, for example, at least one comparator and an associated counter to count pulses exceeding an energy threshold associated with the comparator. Specifically, in a pixelated X-ray inspection system, each pixel may be associated with or include at least one counting unit.

[0047] Based on the data from the counting unit, an image generation unit can generate an image. Advantageously, high-quality images can be obtained.

[0048] The counting unit can provide count rate values ​​determined for multiple energy levels of X-ray photons, for example, by including multiple comparators with associated energy thresholds and counters. Furthermore, the counting unit can provide count rate values ​​based on the X-ray photon flux. Specifically, the count rate can vary with the X-ray photon flux. Additionally, the counting unit can provide spectral curves of the count rate values. Therefore, the counting unit can provide a set of spectral curves of the count rate values ​​as a function of X-ray photon flux versus photon energy, or a set of flux curves of the count rate values ​​as a function of photon energy versus X-ray photon flux (or tube current).

[0049] The aforementioned objective is also achieved by the photon-counting X-ray imaging system described above, wherein the setpoint value of the control device is adjustable, and the counting unit is capable of providing a flux-dependent count rate response, such that the spectral characteristics of the flux-dependent count rate response can be adjusted using the setpoint value of the control device. The setpoint value represents the target point of the statistical value. In other words, the setpoint value is used to influence the spectral characteristics (linear curve) of the flux-dependent count rate response. Specifically, the position of the spectral characteristics can be changed by altering the setpoint value.

[0050] According to another embodiment, the setpoint value of the control device is adjusted to minimize the flux-dependent offset of features in the spectral characteristics. For example, local maxima in the spectrum change their position as a function of flux (e.g., the peak height of the output pulse corresponding to a specific feature line in the input X-ray spectrum can increase at higher fluxes due to accumulation with other concurrent pulses). This flux-dependent correlation can be reduced by selecting an appropriate setpoint value.

[0051] According to another aspect of the invention, a computed tomography apparatus comprising the photon-counting X-ray imaging system as described above can be provided. Therefore, the CT apparatus can also benefit from the advantages of the photon-counting X-ray imaging system of the present invention (e.g., reduced flux correlation with energy scaling, flux-independent contrast, and / or flux-independent Hounsfield units).

[0052] The above objective is also achieved by a signal processing method, the method comprising the following steps:

[0053] - Receive input signals,

[0054] - Determine the output signal from the input signal.

[0055] - Determine statistical values ​​from the output signal.

[0056] - Determine the input signal modification (e.g., offset) that reduces or minimizes the difference between the statistical value and the setpoint value.

[0057] Specifically, the above objective is achieved by using a method for processing signals using an X-ray detection system as described above, the method comprising the following steps:

[0058] - The input signal is received from the detector unit through the input device of the signal processing unit.

[0059] - The output signal is determined from the input signal by a signal processing device.

[0060] - The control unit of the signal processing device determines the statistical value from the output signal.

[0061] - Determine an input signal modification, said input signal modification being used by the control device of the signal processing apparatus to reduce the difference between the statistical value and a setpoint value, and

[0062] - The input signal is modified and applied to the input signal by the control device.

[0063] Typically, this reduction in statistical characteristic differences only becomes visible when the output signal is updated.

[0064] The advantages and further developments of signal processing devices, X-ray detection systems, photon-counting X-ray imaging systems, and computed tomography devices are also applicable to the method of the present invention. The aforementioned functional features of each device can be regarded as corresponding method features.

[0065] In addition, a computer program or computer-readable medium including instructions may be provided, which, when executed by the signal processing apparatus as described above, cause the signal processing apparatus to perform the methods described above. Attached Figure Description

[0066] The invention will now be described in more detail with reference to the accompanying drawings, in which:

[0067] Figure 1 This is an exemplary embodiment of a computed tomography (CT) apparatus;

[0068] Figure 2 Basic implementation of a baseline maintainer;

[0069] Figure 3 Here is an example of how to implement a baseline maintainer, and

[0070] Figure 4 and Figure 5 It is a portion of the differential count rate spectrum of different fluxes of X-ray photons. Detailed Implementation

[0071] The following implementation schemes represent preferred embodiments of the present invention.

[0072] Figure 1 A schematic representation of an advantageous embodiment of the proposed X-ray apparatus as a medical CT (computed tomography) device 1 is shown. The CT apparatus 1 may include an X-ray source 2, a photon-counting X-ray detector including a detector unit 3, and a processing unit 11 as a signal processing device. The signal processing device and the detector unit are shown separately here. However, in an advantageous variation, these are implemented adjacent to each other. In particular, the signal processing device may be constituted by electronic circuitry connected to the detector unit, for example, in the form of an ASIC. The X-ray source 2 and the X-ray detector unit 3 may be arranged opposite each other. The X-ray source 2 may be configured to irradiate the X-ray detector unit 3 with X-rays along the X-ray incident direction. The X-ray detector unit 3 may include a direct-conversion (semiconductor) X-ray detector layer as a sensor layer. For example, the X-ray detector layer may include CdTe, CdZnTe, CdTeSe, CdZnTeSe, CdMnTe, Si, GaAs, or Cr:GaAs as semiconductor materials.

[0073] The CT apparatus 1 may also include a gantry 4 with a rotor 5. An X-ray source 2 and an X-ray detector unit 3 may be arranged on the rotor 5 in a defined configuration, particularly integrated into or attached to the rotor 5. The rotor 5 may be mounted to rotate about a rotation axis 6. The object to be imaged 7 may be mounted on a patient positioning device 8 and may move along the rotation axis 6 through the gantry 4. A processing unit 11 may be used to control the CT apparatus 1 and may include an image generation unit for calculating cross-sectional or volumetric images of the object 7. The processing unit 11 may include a counting unit for counting the pulses generated by the detector unit 3 when detecting X-ray photons. However, in an advantageous variant, the counting unit may also be at least partially included in the electronic circuitry described above, which is connected to and adjacent to the detector unit. An input device 9 (e.g., a keyboard) and an output device 10 (e.g., a screen and / or display) may be connected to the processing unit 11, particularly via signal technology coupling. The input device 9 may advantageously be integrated into the output device 10, for example, in the case of an input display, particularly a resistive and / or capacitive input display. The output device 10 can be designed to display a graphical representation of the counting signal and / or X-ray image dataset.

[0074] The schematic representations included in the accompanying drawings do not depict any scale or proportion.

[0075] Another example (not shown in the figure) may involve a single-plane X-ray system having a C-arm held by a support in the form of a six-axis industrial robot or a multi-joint robot. An X-ray radiation source, such as an X-ray source with an X-ray tube and collimator, is attached to the end of the single-plane X-ray system as an image acquisition unit, and an X-ray image detector as detector unit 3. The implementation of the X-ray diagnostic device does not depend on an industrial robot. A conventional (fixed or movable) C-arm device can also be used.

[0076] In the example described, the patient or technical object to be examined can be positioned on the table of a patient positioning stage within the beam path of the X-ray emitter. A system control unit with a computer for image processing is connected to the X-ray diagnostic apparatus, which receives and processes image signals from the X-ray image detector. The system control unit may include the photon-counting X-ray imaging system and / or signal processing device as described above. The X-ray image can then be viewed on a monitor lamp display. The monitor lamp can be held in place by a ceiling-mounted, longitudinally movable, pivotable, rotatable, and height-adjustable support system having cantilevered arms and lowerable support arms. A counting unit for counting X-ray pulses may also be provided in the system control unit.

[0077] In a particular embodiment, the signal processing device 11 uses a baseline hold (BLH) as a control device with a quantile-based setting to compensate for dark current and achieve a position in the count rate spectrum independent of photon flux. The basis for controlling the BLH is a distribution function that describes the probability that the signal is below a certain energy threshold. Preferably, the BLH continuously determines this percentage of the signal below the baseline (the time below the baseline, TBB) and shifts the signal such that this percentage corresponds to a set quantile value. For example, if a 30% quantile is set, the signal is shifted so that it is 30% below the baseline for that period of time. The baseline itself is typically chosen to correspond to an energy of 0 keV.

[0078] Figure 2 The flowchart illustrates possible signal processing for the signal from X-ray detector unit 3. This signal is referred to herein as the input signal IS. The input signal IS is typically a time-varying pulse signal. The input signal is fed to the positive input of subtractor 12 (or the negative input is fed to the adder). Alternatively, the polarity of the signal can be reversed at multiple points in the circuit without changing the scope of the invention. The designations “positive” / “negative” are used herein to provide a self-consistent interpretation and are not limited to this specific choice of polarity group. The output of subtractor 12 is provided to amplifier 13, which amplifies the signal. The amplified signal can optionally be fed to shaper 14 to shape the pulse of the pulse signal. As a result, an output signal OS is provided at the output of shaper 14 or amplifier 13. Alternatively, the amplifier may have an inherent shaping function, thus eliminating the need for a subsequent shaper, and the output signal OS is provided directly by the amplifier.

[0079] The output signal OS is fed to the baseline hold 15, which generates the compensation current CC. The compensation current CC is provided to the negative input of the subtractor 12. Therefore, the compensation current CC is subtracted from the original input signal IS. The difference signal is provided to the amplifier 13. This means that a feedback loop is provided, in which the baseline hold 15 feeds back the processed output signal OS to the input signal IS.

[0080] The reference hold 15 includes a comparator 16. The comparator 16 compares the output signal OS with a baseline BL, which can be provided externally. Specifically, the baseline BL can be manually set via an interface.

[0081] The output of comparator 16 is fed to weighting unit 17 for quantile-based weighting. The output of weighting unit 17 is input to integrator 18, which integrates the signal. The output of integrator 18 is provided to voltage-to-current converter 19. Voltage-to-current converter 19 generates a compensation current CC based on the integrated signal.

[0082] As described above, the input signal IS is first amplified and optionally pulse-shaped. Comparator 16 of BLH15 compares the amplified / pulse-shaped signal with a reference, i.e., the baseline voltage BL, and outputs only the signal in binary form, regardless of whether the signal is currently above or below the baseline. The two phases are then weighted according to a selected quantile: the phase below the baseline is weighted relative to the phase above the baseline by a factor of (1 - TBB) / TBB. To set a 30% quantile, for example, a factor of 0.7 / 0.3 = 2.333 would be necessary for TBB = 0.3. The weighted signal is finally integrated to obtain a control voltage, which is converted to a compensation current CC by a voltage / current converter 19 and fed to the detector input, for example, by adding it to or subtracting it from the input signal IS.

[0083] A concrete implementation example of the described BLH concept is in Figure 3 As shown in the diagram. For example, a limiting differential amplifier 20 can be used as comparator 16 here. Weighting is performed by a weighting unit 17, which includes two current sources 21 and 22 of different polarities, whose absolute values ​​have a desired ratio of (1 - TBB) / TBB to each other, and are alternately activated according to the output signal of comparator 16. If the output signal OS is below the baseline BL (state "0"), the current source 21 connected to VDD is switched to integrator 18 to feed positive current to integrator 18. In state "0", the other current source 22 connected to ground is decoupled from integrator 18. Otherwise, if the output signal OS is above the baseline BL (state "1"), the current source 22 connected to ground is switched to integrator 18 to feed negative current to integrator 18 (draw current from integrator 18). In state "1", the other current source 21 connected to VDD is decoupled from integrator 18.

[0084] The individual currents are then fed to a differential amplifier 23, which is connected as an integrator, to form an integrator 18. For example, a resistor 24 is used to convert the output voltage of integrator 18 into a compensation current CC.

[0085] If current sources 21 and 22 remain configurable, a variety of quantiles can be selected for control.

[0086] Although the desired quantile is set by the ratio of the two currents, the rate of adjustment can also be set independently of this by selecting appropriate average current I / 2 and integrating capacitor Cintegr, so as to match the requirements of the application.

[0087] Other implementations of the described BLH concept are possible because various circuits can be used. Figure 1The different components are shown. For example, a transistor can also be used as a voltage-to-current converter. In the simplest case, an integrator can consist of only a single capacitor.

[0088] Theoretically, even more complex implementations could be conceived, such as using a time-to-digital converter that converts the durations of the two comparator states into digital values ​​and performs further processing (weighting, integration) in the digital domain.

[0089] This signal processing allows for energy classification of pulse measurements to be as stable and robust as possible over a wide range of possible X-ray fluxes.

[0090] Using a BLH with quantile-based adjustment and its circuit implementation yields various advantages:

[0091] - The signal baseline does not need to be visible, for example, as required by the function of BLR. This means that its use is not limited to applications with low photon flux.

[0092] - In low-flux conditions, the signal baseline is calibrated to the desired reference baseline, regardless of the chosen quantile. This means that the described BLH always behaves very similarly to a BLH with DC or AC coupling behavior in low-flux conditions, and effectively only compensates for dark current.

[0093] At higher photon fluxes, the BLH setting, i.e., the selected quantile, defines the direction and intensity of the shift in the count rate spectrum. Appropriate selection of the quantile (e.g., based on calibration measurements) ensures that the position of the count rate spectrum is virtually independent of the photon flux of a given input signal spectrum. Conventional BLHs with DC or AC coupling behavior cannot achieve this.

[0094] - Due to the position of the count rate spectrum independent of photon flux, BLH is particularly suitable for applications where high or strong fluctuations in photon flux can occur (e.g., medical X-ray imaging).

[0095] - Quantile-based settings are particularly well-suited for simple implementations. This is because the operating mode is based solely on the ratio of two variables (e.g., two current sources), and therefore can be configured very easily and implemented independently of process, voltage, and temperature.

[0096] The BLH's response rate to changes in photon flux can be freely adjusted independently of the selected quantile. This allows it to be easily adapted to the requirements of various applications. For example, in the field of medical X-ray imaging, adjustment can be selected within the measurement interval.

[0097] The possibility of adjustable / configurable quantiles means that the BLH can be adapted to different pulse shapes and amplifier topologies.

[0098] - Selectable quantiles can also be used to achieve a desired flux-dependent offset of the energy scale in the positive or negative direction. This is particularly useful for improving linear behavior (at the expense of energy stability), since higher energy thresholds will later paralyze.

[0099] Therefore, the main advantage of this concept is that it can be used to achieve near-flux-independent stability of the energy scale. For use in computed tomography, this means that the measured contrast and HU values ​​are independent of flux. This is a key requirement for quantitative CT imaging. The aforementioned advantages are... Figure 4 and Figure 5 As shown in the image.

[0100] Figure 4 Cross sections of differential count rate spectra for different fluxes of X-ray photons are shown. Higher flux results in higher counts (i.e., the corresponding spectra). The spectra (also referred to herein as spectral characteristics) have two exemplary spectral features: a first local maxima 25 and a second local maxima 26. Higher flux results in more broadening or narrowing of the corresponding local maxima 25, 26, and a further leftward shift towards lower energies. However, a rightward shift towards higher energies is also possible. These shifts are readily identifiable relative to the vertical dashed lines 27 and 28. This shift can be compared to the fact that the color of an image changes with increasing brightness. However, this is undesirable.

[0101] like Figure 5 As shown, the method of the present invention can stabilize spectral characteristics. For example, even with increased flux, local maxima 25 and 26 remain constant on vertical lines 27 and 28. This can be achieved by adjusting the quantiles individually. The spectrum or spectral characteristics do not show an energy correlation with flux.

[0102] Simultaneously, it achieves dynamic and automatic adaptation to local and temporal variations in leakage current for each pixel. On the one hand, this eliminates the need for intermediate leakage current calibration measurements. On the other hand, it results in a detector system robust to variations in sensor leakage current. This achieves the spatial and temporal stability of the sensor signal required for CT imaging.

[0103] Therefore, the above concepts represent a favorable means to make photon counting imaging, especially photon counting CT, more reliable.

Claims

1. An X-ray detection system, comprising a detector unit (3) and a signal processing device (11), wherein the detector unit (3) is used to generate an input signal (IS) based on detected X-ray photons, and the signal processing device (11) is used to process the input signal, wherein, The signal processing device (11) includes: - An input device for receiving the input signal (IS). - A processing apparatus for generating an output signal (OS) from the input signal (IS), - A control device for influencing the input signal (IS), - The control device is capable of: Determine the statistical values ​​of the output signal (OS). The statistical value is compared with a predetermined setpoint value, and The input signal is affected, causing the difference between the statistical value in the output signal and the setpoint value to decrease.

2. The X-ray detection system according to claim 1, wherein, The statistical value is the quantile or duty cycle of the output signal (OS).

3. The X-ray detection system according to claim 2, wherein, The input signal (IS) is a continuous signal, and the quantile refers to the proportion of time the output signal is below a predetermined baseline.

4. The X-ray detection system according to any one of the preceding claims, wherein, The input signal (IS) includes a pulse signal.

5. The X-ray detection system according to any one of the preceding claims, wherein the processing device includes a charge-sensitive amplifier or a transimpedance amplifier for amplifying the input signal (IS).

6. The X-ray detection system according to claim 5, wherein, The control device can add a compensation current (CC) to the input signal (IS), the compensation current (CC) being based on the output signal of the corresponding amplifier.

7. The X-ray detection system according to claim 6, wherein, The compensation current is generated by the baseline hold circuit (15) or the baseline restorer circuit based on the output signal of the respective amplifier.

8. The X-ray detection system according to claim 7, wherein, The baseline hold circuit (15) includes an integrator circuit (18) and two adjustable current sources (21, 22) of opposite polarity connected to the input of the integrator. The output from the integrator (18) is the basis of the compensation current (CC). One of the adjustable current sources provides current only when the output signal (OS) is below a predetermined baseline (BL), and the other adjustable current source provides current only when the output signal (OS) is above the predetermined baseline (BL).

9. The X-ray detection system according to claims 3 and 8, wherein, The current intensity of one of the adjustable current sources with opposite polarities is proportional to the time proportion during which the output signal (OS) is above the predetermined baseline (BL), and the current intensity of the other adjustable current source is proportional to the time proportion during which the output signal (OS) is below the predetermined baseline (BL).

10. A photon-counting X-ray imaging system, comprising: - An X-ray detection system according to any one of the preceding claims, wherein the input signal (IS) includes a corresponding pulse signal, and - A counting unit for counting pulses of the output signal (OS) that exceed a predetermined energy threshold.

11. The photon-counting X-ray imaging system according to claim 10, wherein, The setpoint value of the control device is adjustable, and the counting unit is capable of providing a flux-dependent count rate response, such that the spectral characteristics of the flux-dependent count rate response are adjustable using the setpoint value.

12. The photon-counting X-ray imaging system according to claim 11, wherein, The setpoint value of the control device is adjusted to minimize the flux-related offset of the features in the spectral characteristics.

13. A computed tomography apparatus (1) comprising a photon counting X-ray imaging system according to any one of claims 10 to 12.

14. A method for processing signals using an X-ray detection system according to any one of claims 1 to 9, the method comprising the following steps: - Receive input signal (IS) from the detector unit (3). - Determine the output signal (OS) from the input signal (IS). - Determine statistical values ​​from the output signal (OS). - Determine the input signal modification (CC) to reduce the difference between the statistical value and the setpoint value, and - Apply the input signal modification (CC) to the input signal (IS).

Images