PHOTON COUNTING

Abstract

A method for photon counting for pixels in a pixelated detector is disclosed, wherein for each of the pixels, one or more neighbouring pixels are defined. The method comprises receiving a charge in one or more of the pixels and comparing for each of the pixels the charge with a trigger threshold. If the charge in a pixel is above the trigger threshold, the charge is registered in the pixel after a registration delay, wherein the registration delay is dependent on the level of the charge received in the pixel in such a way that a registration delay decreases with increasing charge. A counter for a pixel is incremented when the charge is registered and an increment of a counter of the neighbouring pixels is inhibited. Pixelated semiconductor detectors are also disclosed.

Claims

1. A method for photon counting for pixels in a pixelated detector, wherein for each of the pixels, one or more neighbouring pixels are defined, the method comprising: receiving a charge in one or more of the pixels; for each of the pixels comparing the charge with a trigger threshold, and if the charge in the pixel is above the trigger threshold further comprising: registering the charge in the pixel after a registration delay, wherein the registration delay is dependent on the level of the charge received in the pixel in such a way that a registration delay decreases with increasing charge; incrementing a counter for the pixel when the charge is registered; and inhibiting an increment of a counter of the neighbouring pixels.

2. The method of claim 1, further comprising collecting the charge of the neighbouring pixels.

3. The method of claim 1, wherein registering the charge in the pixels with a registration delay comprises for each of the pixels discharging a capacitor at a fixed rate from a maximum level to a reference level after a set delay, wherein the reference level is dependent on the charge received in the corresponding pixel.

4. The method of claim 3, wherein the reference level is equal to or is directly related to a peak value of a pulse registered in the corresponding pixel.

5. The method of claim 1, wherein for each of the pixels a plurality of trigger thresholds with different values are defined, each of the trigger thresholds having a different predefined trigger delay, in such a way that the predefined trigger delay decreases with increasing value of the threshold, and wherein comparing for each of the pixels the charge with a trigger threshold comprises comparing the received charge with each of the trigger thresholds; and if the charge in a pixel is above one or more trigger thresholds registering the charge in the pixels with a delay comprises registering the trigger with the shortest delay.

6. A pixelated semiconductor detector having a plurality of pixels, and a read-out circuit, wherein the read-out circuit is configured for each of the pixels, to register a charge in the pixels with a registration delay, wherein the registration delay is dependent on the level of the charge received in the pixel in such a way that a registration delay decreases with increasing charge; and the read-out circuit comprises a counter for each of the pixels for counting a registration of the charge, and a communication module for communicating the registration to neighbouring pixels, and wherein the counters are configured to ignore a registration when the communication of a registration is received from a neighbouring pixel.

7. The pixelated detector of claim 6, wherein the read-out circuit further is configured to determine the charge collected by a pixel in an event.

8. The pixelated detector of claim 7, wherein the read-out circuit comprises for each of the pixels a summing module for adding the charge collected by neighbouring pixels in the event.

9. The pixelated detector of claim 6, wherein the read-out circuit comprises a capacitor configured to discharge after a set delay from a maximum voltage to a reference voltage, wherein the reference voltage is dependent on the charge collected by the corresponding pixel, and wherein the charge is registered once the capacitor reaches the reference voltage.

10. The pixelated detector of claim 9, wherein the reference voltage is equal to or is directly related to a peak value of a pulse registered in the corresponding pixel.

11. The pixelated detector of claim 9, wherein the discharge rate and the set delay are independent from the charge collected by the corresponding pixel.

12. The pixelated detector of claim 6, wherein the read-out circuit comprises for each of the pixels, a plurality of triggers with thresholds with different values, each of the trigger thresholds having a different predefined trigger delay, in such a way that the predefined trigger delay decreases with increasing value of the threshold, and wherein the read-out circuit is configured to compare the received charge with each of the trigger thresholds; and if the charge in a pixel is above one or more trigger thresholds to register the trigger with the shortest delay.

13. The pixelated detector of claim 6, wherein the read-out circuit comprises a pulse shaper for each of the pixels.

14. An imaging device comprising a pixelated detector according to claim 6.

15. The pixelated detector of claim 9, wherein the capacitor is configured to discharge only if the received charge is above a predetermined threshold.

16. A pixelated semiconductor detector having a plurality of pixels, and a read-out circuit, wherein the read-out circuit is configured to register a charge of an event in the pixels with a registration delay, wherein the registration delay decreases with increasing charge received in the pixels; to define the pixel for which the charge of the event is registered first as a master pixel for the event, and to increment a counter for the master pixel, and to inhibit a counter for pixels neighboring the master pixel for the event.

17. The pixelated semiconductor detector according to claim 16, the read-out circuit further is configured to determine the charge collected by the master pixel in the event.

18. The pixelated semiconductor according to claim 17, wherein the read-out circuit is furthermore configured to determine the charge collected by the neighboring pixels and to assign the charge collected by the neighboring pixels to the master pixel for the event.

19. The pixelated semiconductor detector according to claim 16, wherein the read-out circuit comprises for each pixel a capacitor configured to discharge if a charge above a predefined threshold is received, and wherein the charge is registered once a capacitor reaches a reference voltage.

20. The pixelated semiconductor detector according to claim 19, wherein the reference voltage is dependent on the charge collected by the corresponding pixel.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Non-limiting examples of the present disclosure will be described in the following, with reference to the appended drawings, in which:

[0033] FIG. 1 schematically illustrates a prior art method and system for photon counting;

[0034] FIG. 2a schematically illustrates a pixelated detector according to an example;

[0035] FIG. 2b schematically illustrates some characteristics of a pulse as may be received in a detector;

[0036] FIGS. 3a-3e illustrate an example of a method and a system for photon counting according to an implementation;

[0037] FIGS. 4a-4b illustrate an example of a method for photon counting according to another implementation; and

[0038] FIGS. 5a and 5b schematically illustrate an example of a system for photon counting according to an implementation.

DETAILED DESCRIPTION

[0039] FIG. 1 schematically illustrates a prior art method and system for photon counting. Reference sign 20 refers to a pixel of a pixelated electrode with corresponding detector material. X-ray photons 21 can impinge on a phototransducer such as a photodiode 22. This can create a small charge packet of electrons, which can be shaped to a pulse 25 by a pulse shaper circuit 23. A comparator 24 or “discriminator” compares the generated pulses 25 to a reference value or trigger level.

[0040] Every time the amplitude of the pulses 25 exceeds the trigger level, a first value, for example “high” or a digital one, is assigned to the output signal of the comparator 24, and every time the pulses are below the trigger level, a second value, for example “low” or digital zero, is assigned to the output signal of the comparator 24.

[0041] This way, a binary signal is generated, comprising a pulse train 26. The pulses in the pulse train 26 are then counted by means of a digital counter 27 in order to provide a count value for the impinging incident photons 21.

[0042] As discussed before, whenever a charge is shared between pixels, a single event can be counted more than once. If the pulse of more than one pixel reaches above the trigger, for each of the pixels a digital one will be assigned.

[0043] FIG. 2a schematically illustrates a pixelated detector according to an example. A simplified example having an array or matrix of 5×5 pixels is shown. In accordance with this example, each of the pixels is provided with a communication module that enables communication between neighbouring pixels. For example, to effectively avoid double counting in pixel 23 (and neighbouring pixels), pixel 23 is enabled to communicate with pixels 12, 13, 14, 22, 24, 32, 33 and 34. Pixel 51 is enabled in a similar manner to communicate with pixels 41, 42 and 52.

[0044] If in a specific event, pixel 23 is the master pixel, pixel 23 can inhibit the counter of each of the slave pixels 12, 13, 14, 22, 24, 32, 33 and 34. Also, the master pixel can request each of the slave pixels to communicate the charge they received so that the total charge can be assigned to the master pixel.

[0045] In most examples, neighbouring pixels may be defined herein as a pixel sharing a portion of a border with another pixel (such as e.g. pixels 23 and 24), or at least a point along the border (such as e.g. pixels 12 and 23). In further examples, neighbouring pixels may be defined as pixels between which charge sharing can potentially occur. The pixels between which charge sharing could potentially occur could be calculated by taking into account the drift potential (depending on bias voltage, detector thickness and pixel size, photon energy levels).

[0046] FIG. 2b schematically illustrates some characteristics of a pulse as may be received in a detector. The illustrated pulse may be output of a pulse shaper.

[0047] It is known that the shape of a pulse may vary. The shape of the pulse can be characterised by a peak voltage (V.sub.pk), the rise time (ΔT.sub.r), i.e. the time between the moment the signal passes the trigger threshold and the moment of peak voltage, and fall time (ΔT.sub.f), i.e. the time between the moment of peak voltage and the moment the signal passes through the trigger threshold again.

[0048] In case of charge sharing, the pulses received by various pixels may generally be similar in that the rise time and fall time will be substantially the same. The peak voltage will however vary. The pixel that receives the highest charge will have the highest peak voltage.

[0049] FIGS. 3a-3e illustrate an example of a method and a system for photon counting according to an implementation.

[0050] As proposed in various examples disclosed herein, a key to the solution for the charge sharing problem is that it can be determined which pixel has the maximum charge deposition. Then that pixel can be assigned to be the master pixel that should have its counter incremented and, if needed, the energy digitized and stored. The neighbouring pixel(s) that may receive part of the charge, should not increment their counters. In some examples, the neighbouring pixels may pass the charge that they have collected to the master pixel for further processing.

[0051] FIGS. 3a and 3b illustrates a first principle upon which the assignment of a master pixel can be based. In each of the pixels, wherein a charge above the threshold is received, the discharge of a capacitor is triggered. However, this discharge only occurs after a set delay ΔT.sub.delay. The set delay may be the same for all pixels.

[0052] The difference in discharge between pixels is caused by setting a reference level at which discharge is considered completed to the peak level of the received signal (taking into account the gain of the pixel electronics). A peak hold circuit may be used for this purpose.

[0053] A higher peak (V.sub.pk) means the energy deposited in that pixel is higher and that the discharge will be completed sooner. This is illustrated in FIG. 3b: since V.sub.pk2 is higher (and V.sub.max, as well as the discharge rate, and ΔT.sub.delay is the same in both cases), the discharge of pixel 2 is completed sooner and this pixel can thus be assigned to be the master pixel. The completion of the discharge causes a second trigger.

[0054] The set delay for starting the discharge may preferably be linked to the rise time, such that it can be ensured that the correct peak voltage can be determined before discharge.

[0055] FIGS. 3c and 3d schematically illustrate mechanisms of the first and second triggers. When the signal reaches the trigger threshold, (after a set delay), switch 1 is flipped, and switch 2 is closed. This causes the discharge of the capacitor through a resistance. During the discharge, the voltage level of the capacitor may be input for a second comparator or discriminator, illustrated in FIG. 3d.

[0056] The comparator can compare the peak voltage level registered in a pixel with the voltage of the capacitor. As the capacitor discharges, this voltage decreases. At the moment the voltage of the capacitor equals the peak voltage, the discharge is completed. This moment can be communicated to neighbouring pixels, such that their counter is not increased.

[0057] Although in this example, the reference level is set equal to the peak voltage level of the pulse, it should be clear that alternatives are possible as long as the reference level correctly varies with the energy deposited in the pixel, i.e. the reference level should be higher for higher energies.

[0058] In some examples, the charge collected by neighbouring pixels may then be passed on to the master pixel, such that the total charge can be determined.

[0059] FIG. 3e schematically illustrates a system and method for photon counting in accordance with this implementation. As explained before with reference to FIG. 1, photons 21 may impinge on a photodiode 22. The signal may be passed through a shaper 23. The first trigger may be caused as before, i.e. reaching a trigger level. Comparator or discriminator 24 may be used as described before with reference to FIG. 1. However, as a consequence of this trigger, after a certain delay the switches S1 and S2 of trigger mechanism 30 reach the positions indicated in FIG. 3e.

[0060] The voltage level of the capacitor is used as input for the second trigger mechanism 40. The other input may be the peak voltage and may be provided by shaper 23. The output of the second trigger may be binary pulses 26 which may be counted by counter 27. Communication module 50 may communicate the event to neighbouring pixels such that double counting can be avoided.

[0061] In normal operation, the user knows the range of the X-ray photons that will be emitted from the X-ray tube or natural source. Based on this knowledge the user can set the maximum dynamic range of the pixel. If the X-ray tube emits a spectrum of X-ray photons from 20 keV to 100 keV and if one assumes that the gain of the pixel electronic is 10 mV/keV, then the maximum expected pulse height is 1000 mV. Let us assume that the pulse shape of the pixel has a rise-time of 100 nsec, and a fall time of 400 nsec, and the maximum jitter time of the discriminator is around 25 nsec. In this example the ΔT.sub.w (time-walk)=rise-time+jitter time=125 nsec.

[0062] For this example one may set the minimum pixel threshold at 20 mV above the base line to make sure that the level is about 5 or 6 standard deviations above the random noise level.

[0063] Let us assume that a photon of 80 keV has impacted close to the edge of the pixel in which 75% of the charge was deposited and that 25% of the charge is deposited in the adjacent pixel. Each pixel, in this example, can have a reference-capacitor that is already charged to 1000 mV (equivalent to 100 keV of the maximum energy level of the X-ray photon. The maximum level can be adjustable).

[0064] With a signal charge equivalent to 60 keV in one pixel and 20 keV in the adjacent pixel, both pixels will trigger because the threshold of the discriminator was set for 2 keV (20 mV). The pixel trigger will initiate, after a set delay of peak time (100 nsec in this example), the discharge of the reference capacitor at a fixed rate of RC=50 nsec (decay constant). I.e. the discharge of the capacitor will start 125 nsec after the peak (and 225 nsec after reaching the trigger value).

[0065] In the first pixel, one can say that the discriminator-2, comparing the level from the discharging reference-capacitor and peak hold signal of the 60 keV is at around 55 nsec ahead in time of the trigger coming from the discriminator-2 in the second pixel by comparing the level of the discharging reference-capacitor and the peak signal of 20 keV. The trigger output from discriminator-2, in each pixel, will open switch 2 to stop the discharge of reference-capacitor. The registration of the charge will thus start to happen 125 nsec after the peak moment (and the peaking time will be substantially the same for all pixels). The highest charge will be registered 150 nsec after peaking.

[0066] In this example, the trigger from discriminator-2 of the pixel with 60 keV energy deposition, will be communicated first to all surrounding triggered pixels to indicate two things, namely to inhibit their counters, and to request the peak hold signal to be communicated to the master pixel (for further processing specially in spectroscopy information is needed). The energy collectively registered in all the triggered pixels can be measured in two ways: in an analogue scheme by summing up the values of the voltage of the reference-capacitor in each triggered pixel, or alternatively in a digital scheme by measuring the time for the ref-capacitor to discharge from V.sub.max to V.sub.pk in every pixel.

[0067] FIGS. 4a and 4b schematically illustrate an alternative method and system for photon counting. As illustrated in FIG. 5a, a plurality of threshold levels may be defined Thr.sub.1, Thr.sub.2, Thr.sub.3, . . . Thr.sub.n. As soon, as the signal reaches a threshold level, a trigger occurs. A single signal can thus cause n triggers as illustrated in FIG. 4a.

[0068] FIG. 4b schematically illustrates how these different thresholds can be used for effective photon counting while avoiding double counting. Each of the thresholds levels may have different delays. The delay may be set to be shorter for triggers corresponding to higher energy levels.

[0069] In the example of FIG. 4b, the registration delay is related to the rise time (such that the peak voltage can be determined reliably), to which a delay is added of a constant (maximum time-walk) and a delay that changes for each of the triggers, e.g. one rise time for the highest threshold, second rise times for the threshold below that, and thus “n” times the rise time for the first threshold.

[0070] In different examples, different schemes for delays may be used. For example, instead of adding a single rise time ΔT.sub.r for each of the consecutive trigger thresholds, a constant ΔT.sub.x=ΔT.sub.r+20 nsec could be used. In this case, the delay for the highest threshold could be a walk time +ΔT.sub.x, for the second highest threshold the delay would be the walk time +2 ΔT.sub.x etc.

[0071] In yet further examples, a different constant that is independent of the rise time could be used, e.g. 50 nsec. The delay for the highest threshold would be the time-walk+50 nsec, the delay for the second highest threshold would be the time-walk+100 nsec etc.

[0072] A schematic arrangement showing the different delays and counters is shown in FIG. 5a. The output of a pulse shaper 23 may be fed to a plurality of discriminators 24, 34, 44 having different threshold values. The delay for registration or triggering is different for each of the discriminators. Of course, a discriminator will not register anything, i.e. there will be no trigger if the peak value of the pulse is below the corresponding trigger threshold.

[0073] The set delay for each of the triggers is different. In this example, the delay for each of the triggers may comprise a constant (k) and a part that varies for each of the triggers. If n trigger thresholds are defined, the delay for the highest threshold may be set equal to constant k. Constant k may be (as in the example of FIG. 4b) equal to rise time+walk time. The varying part of the delay may be made dependent on the rise time. For the second highest threshold, this part may be equal to one rise time. For the threshold below that time, it may be equal to two rise times. For the lowest threshold, it may thus be equal to (n−1) rise times. The registration delay for the lowest threshold may thus be equal to k+(n−1) ΔT.sub.r.

[0074] FIG. 5b schematically illustrates how the definitive count can be influenced by the information received from neighbouring pixels. A counter is incremented only if it is not inhibited by a neighbouring pixel or by a comparator belonging to a different threshold.

[0075] As illustrated in FIG. 4b, this ensures that the register will occur first for the highest level. Let's imagine that the minimum threshold is set to 10 keV and the step between two thresholds is set to 15 keV. Let us also imagine that there is a total 8 thresholds covering the range from 10 keV to 130 keV for each of the pixels. The thresholds are thus set at 10, 25, 40, 55, 70, 85, 100, 115, and 130 keV.

[0076] If a photon has deposited a total energy of 80 keV in one pixel, then the triggers will occur for the levels of 10, 25, 40, 55 and 70 keV. The trigger corresponding to 70 keV will occur first. The counter of capacitor 5 will be incremented by 1.

[0077] If the charge is split between two pixels such as 45 keV and 35 keV, then in the first pixel, triggers corresponding to 10, 25 and 40 keV will occur. In the second pixel, the triggers corresponding to 10, and 25 keV will occur. However, the third trigger of the first pixel will be registered first. The counter of trigger 3 will be incremented by 1. At this moment, its registration is communicated to neighbouring pixels, such that the counter of trigger 2 of pixel 2 is not incremented.

[0078] This arrangement makes it particularly easy to determine (approximately) the total charge deposited. In this case, the estimated charge is between 40 keV and 55 keV (of pixel 1)+between 25 keV and 40 keV (of pixel 2). The charge may thus be estimated at between 65 keV and 95 keV, or at 80 keV as a mean value.

[0079] Although only a number of examples have been disclosed herein, other alternatives, modifications, uses and/or equivalents thereof are possible. Furthermore, all possible combinations of the described examples are also covered. Thus, the scope of the present disclosure should not be limited by particular examples, but should be determined only by a fair reading of the claims that follow.