Method to Reduce the Number of Signals to be Read Out in a Detector

Abstract

The present invention refers to a method to reduce the number of signals to be read out in a detector characterized in that it comprises using a lower photosensor granularity at least at the center obtaining a granularity degree at the photosensor that is edge dependent, wherein the granularity at the detector corners can be higher than at any other detector zone or wherein the granularity at the detector edges can be higher than at any other detector zone, and wherein the number of signals to be read out is reduced by joining signals in at least the center zone of the detector, or is reduced by using different photosensor elements at the center of the photosensor with regard to the remaining photosensor zones, and its use in nuclear medicine imaging techniques.

Claims

1. A method to reduce the number of signals to be read out in a detector having an array of sensor elements with a plurality of rows and a plurality of columns, comprising the steps of: configuring at least a portion of the sensor elements in a center zone of the array to form a lower photosensor granularity than at a periphery of the array obtaining a granularity degree at the photosensor that is edge dependent.

2. The method according to claim 1, wherein the granularity at the detector corners is higher than at any other detector zone.

3. The method according to claim 1, wherein the granularity at the detector edges is higher than at any other detector zone.

4. The method according to claim 1, wherein the lower photosensor granularity is achieved by joining signals from the portion of the sensor elements in at least the center zone of the detector.

5. The method according to claim 1, wherein the lower photosensor granularity is achieved by using different sensor elements at the center zone of the photosensor with regard to the remaining photosensor zones.

6. The method according to claim 4, wherein the number of signals to be read out is reduced by joining signals: at the detector center, or at the detector center and at the detector laterals, or at the detector center, at the detector laterals and at the detector corners.

7. The method according to claim 5, wherein lower photosensor granularity is achieved by using smaller photosensor elements at the detector corners than at the detector center and detector laterals.

8. The method according to claim 5, wherein the lower photosensor granularity is achieved by using smaller photosensor elements at the detector corners, larger photosensor elements at the detector laterals and even larger photosensor elements at the detector center.

9. The method according to claim 5, wherein the-lower photosensor granularity is achieved by using photosensor elements of different shape within the detector.

10. The method according to claim 1, wherein the lower photosensor granularity is achieved by: joining signals in at least the center zone of the detector and by using different sensor elements at the center zone of the photosensor with regard to the remaining photosensor zones.

11. The method according to claim 1 further comprising the step of applying a projection readout by summing all the signals for each row and all the signals for each column giving rise to an additional reduction of the photosensor granularity.

12. The method according to claim 11, wherein: a lower granularity in at least a portion of the sensor elements in a center zone of the array is achieved by joining signals from the portion of the sensor elements in at least the center zone of the detector and the application of the projection readout by summing all the signals for each row and all the signals for each column is carried out previous to the step of joining signals from the portion of the sensor elements in at least the center zone of the detector.

13. The method according to claim 1, wherein the detector is a radiation detector.

14. The method according to claim 1, wherein the detector is a gamma radiation detector.

15. A method for reconstructing the impinging position of a gamma ray in a gamma radiation detector, the method comprising the steps of: detecting the radiation coming from a radiation source by means of at least a-one detector module in the gamma radiation detector, wherein the at least one detector module includes a photosensor array having multiple sensor elements arranged in a plurality of rows and a plurality of columns; measuring radiation signals at the photosensor array; and configuring at least a portion of the sensor elements in a center zone of the array to form a lower photosensor granularity than at a periphery of the array obtaining a granularity degree at the photosensor that is edge dependent.

16. The method according to claim 15, wherein the granularity at the detector corners is higher than at any other detector zone.

17. The method according to claim 15, wherein the granularity at the detector edges is higher than at any other detector zone.

18. The method according to claim 15, wherein the photosensor granularity is reduced by at least one of: joining signals, and using different photosensor elements at the center of the photosensor with regard to the remaining photosensor zones.

19. The method according to claim 18, wherein the photosensor granularity is reduced by using at least one of: smaller photosensor elements at the detector corners than at the detector center and detector laterals, and smaller photosensor elements at the detector corners, larger photosensor elements at the detector laterals and even larger photosensor elements at the detector center.

20. The method according to claim 18, wherein the photosensor granularity is reduced by using photosensor elements of different shape within the detector.

21. A detector module to reduce the number of signals required to be read out during nuclear medical imaging, the detector module comprising: a radiation detector including an array of photosensor elements, wherein the array of photosensor elements is arranged in a plurality of rows and a plurality of columns and wherein the photosensor elements are configured with a larger active area at a center of the radiation detector and smaller active area at an edge of the radiation detector.

22. The detector module according to claim 21, wherein the larger active area at the center is obtained by at least one of: summing the active area of two or more photosensor elements, and using different photosensor elements within the photosensor array.

23. The detector module according to claim 21, wherein photosensor elements located at the photosensor array center are merged into groups of at least two, and photosensor elements located at a corner and the edge of the photosensor array are not merged, or are merged in a smaller degree than at the center.

24. The detector module according to claim 21, wherein photosensor elements located at the photosensor array center and the edges are merged into groups of at least two, and photosensor elements located at a corner of the photosensor array are not merged.

25. The detector module according to claim 21, wherein the larger active area at the center is obtained by using photosensor elements of different shape within the photosensor array.

26. The detector module according to claim 21, comprising a scintillation block selected from one of a monolithic crystal and a pixelated crystal.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0112] FIG. 1 shows the maximal readout channels for two examples of photosensor arrays, left 1616 channels read out individually, and right 1212 channels read out individually.

[0113] FIG. 2 shows an example of the light distribution in a monolithic crystal when reading out 1212 channels. The grey scale shows SiPMs with more or less scintillation photons collected.

[0114] FIG. 3 shows the embodiments of the alternative one of the method, where SiPM signals of nearby photosensors are combined to provide a single output. A higher number of photosensor elements are combined at the center (here 4 into 1) in this example, and less at the laterals (2 into 1). Corner photosensors are not combined. The figure also shows the meaning of corner, lateral and center. In this example a reduction from 144 photosensor elements (maximal number of signals) is reduced to 64.

[0115] FIG. 4 shows the embodiment in which photosensors of different size are combined, following the so-called alternative two. The example also reduces from the 1212 maximal number of outputs to 64, combining for instance, 44 mm, 84 mm and 88 mm photosensors. Larger at the center that at the corners, in order to keep the best sampling where the scintillation light is truncated in the case of using monolithic crystals.

[0116] FIG. 5 does not show one of the embodiments, but it represents the so-called projection readout for the 1212 SiPM array example. Therefore, here 12+12 signals would be readout out. The 3D light distribution in the map is projected onto the X and Y axes.

[0117] FIG. 6 shows the third alternative, based on the 1212 SiPM photosensor matrix. In this figure we have used the second method alternative first, and then the projected readout electronics. Therefore, we reduced from 144 signals to only 16.

[0118] FIG. 7 shows again the representation depicted in FIG. 6 but with the projected light distributions composed by 8 signals from the rows and 8 from the columns.

[0119] FIG. 8 shows the result of using an embodiment like the one shown in FIG. 1, where it is possible to resolve well impacts across the whole crystal volume. This was done for a 505015 mm thick LYSO-type crystal. A collimated source was moved from one side of the crystal to the other using the reduction readout and a photosensor array of 1212 elements. The plot shows the measured spatial resolution (measured through the FWHM of the distribution) as a function of the beam position.

[0120] FIG. 9 shows an example of performance of one of the embodiments of the invention and comparison with alternative approaches. Left, results obtained with the described invention using 1212 photosensor and the reduction readout. Center, using a readout as proposed by Popov or Proffitt. Right, an array of 88 elements with larger active area and constant larger pitch. It is observed that left and center provide very similar results, (but left with a 30% of signal number reduction (what is a substantial advantage over the prior art), whereas right shows the lack of resolving power at the left edge due to the wider sampling there. Bottom, profiles of the central row of sources for the three cases. Here, it is observed how the source closest to the left edge (2.5 mm from crystal edge) can be resolved for the standard project readout (12+12 signals), but also for the proposed third alternative invention with only 8+8 signals without appreciable deterioration of the performance at any detector location. However, the standard projection readout with 8+8 does not solve these sources near the edge.

[0121] FIG. 10 shows a comparison between two approaches based on the third alternative, the one described in FIG. 9 (1), and another on which the photosensor elements are first merged (first alternative) and later projected (2). The plot on the right shows the measured spatial resolution as a function of the DOI layer, where DOI1 contain impacts that happen in the first 5 mm entrance of the crystal, DOI2 is the middle layer, and DOI3 the 5 mm closer to the photosensor.

[0122] FIG. 11 shows an example for one of the embodiments, as shown in FIG. 1, but using a crystal array instead of a monolithic block. Also for these types of scintillators, the invention offers an advantage over conventional methods, due to the reduction of signals without compromising the performance. On the left side, a row of pixels is shown for two acrylic windows thickness placed in between the crystal array and the photosensor (1.7 mm and 2.5 mm). Right, flood map for a 3232 crystal matrix (1.6 mm pitch) and its energy spectra.

[0123] With the information given in the references 1 and 2, mentioned above, that explain how to obtain the 3D photon impact position, especially in a monolithic crystal, and the description, the method of the invention can be put into practice.

[0124] An illustrative example is described hereinbelow.

EXAMPLE

Experimental Set-Up

[0125] An experiment with two identical detector blocks, working in coincidence, has been designed. Two LYSO scintillation blocks (505015 mm.sup.3) with specific surface polishing and treatments were used.

[0126] Each detector block also includes an array of SiPMs as depicted in FIG. 1 right. The photosensor matrix is composed of 1212 SiPMs photosensors with 33 mm.sup.2 active area each, and a pitch of 4.2 mm. Two embodiments have been studies. In the first, instead to reading all photosensors individually (144 signals), each SiPM has been connected to the aforementioned projection readout circuitry that provides scintillation light profiles in the X and Y axes. The readout electronics according to third alternative embodiment reduces the 12 signals for each projection to only 8, resulting on a total of 8+8 outputs per photon impact. This reduction is based on keeping the 2 lateral row and columns, but merging the 8 central ones from 2 to 1. In the second, we have followed another of the embodiments described in the third alternative. Here, we have first applied the first alternative in this 1212 SiPM array reducing from 144 signals to only 64, and later project, resulting on 8+8 signals.

[0127] The two approaches still preserves the good sampling of the light at the crystal edges and, thus, a good characterization of the effect produced by the truncation of the scintillation light therein. The proposed readout scheme makes it possible to reduce the total number of signals of a detector of 144 photosensors to only 16. The performance of the proposed detector block module including the projection reduction readout system, has been compared to a standard 88 SiPMs photosensors with a 66 mm.sup.2 active area, and a pitch of 6.33 mm.

[0128] The performance of the detector block described above was evaluated by means of the spatial, energy and DOI resolutions. Perpendicular to entrance face and lateral incidence measurements to the crystal were carried out. The module under study was irradiated with an array of 1111 .sup.22Na sources, 1 mm in diameter and 1 mm height each (4.6 mm pitch), placed in front of a Tungsten collimator (24 mm thick, 1.2 mm diameter holes), which was in contact with the crystal. The centroids of the measured scintillation light distributions, through the projection readout circuit, are calculated using the center of gravity methods. The photon impact DOI is estimated by the ratio of the sum of all 8 signals (photon energy, E) to the maximum signal value (E/Imax). During the data processing, each detector area is subdivided in 600600 virtual pixels, and a software collimation was applied. An energy window of 15% at the 511 keV peak was also applied in the data analysis. The collimation aperture must guarantee the best compromise between the detector spatial resolution and the measurement statistics of the analyzed measurement.

[0129] The measured detector spatial resolution was evaluated using the imaged 1111 .sup.22Na collimated sources, as shown in FIG. 11 left. We calculated the centroid of each source in channels, as shown in the profile of the top panel in FIG. 11 right, using multi-Gaussian distributions. After the calibration to metric units, it is possible to obtain the measured detector spatial resolution for each source as the FWHM. The spatial resolution for each source was calculated as the average of the X and Y projections, and as a function of the DOI layer. The top right panel in FIG. 11 shows the profile of the central row of sources calibrated into mm, the 11 sources can be well distinguished. Right-bottom, DOI distribution of a source placed at 15 mm from the edge. The energy resolution is determined as FWHM/E.sub.centroid. This was done for each selected DOI range.

[0130] Traditionally, in PET systems based on monolithic scintillators, events close to the crystal edge are hardly considered as good data for the image reconstruction due to its poorer characterization. In order to study how close to the crystal edge can events be properly characterized, a second set of experiments was carried out using normal incidence by means of a pinhole Tungsten collimator with 30 mm thickness and with a drilled hole of 2 mm diameter together with a small size .sup.22Na source. In these experiments the source was displaced in small steps of 0.5 mm across the entire X axis of the crystal.

[0131] To provide accurate DOI resolution values, lateral incidence measurements were carried out in steps of 1 mm also using the pinhole collimator. ROIs of 44 mm.sup.2 were carried out at step distances of 5 mm from the crystal edge, reaching the crystal center. This allows one characterization of the DOI in the whole volume of the scintillator.

[0132] Using the impact DOI information for each impact, we have split the data into three DOI regions, namely DOI1 (entrance), DOI2 and DOI3 (exit), which correspond to crystal depths of 15-10 mm, 10-5 mm and 5-0 mm (near the photosensor), respectively. This allows obtaining three different flood maps.

Results

[0133] An overall good performance including, depth of interaction information has been achieved for both embodiments namely 1) projecting first and later reducing, or 2) merging signals and later projecting. We achieved on both a spatial resolution of about 1.9 mm FWHM for the whole scintillation volume, with an average energy resolution of 13% FWHM and a photon depth of interaction resolution (FWHM) of 3.7 mm.

[0134] The main advantages of the detector block according to the invention is a new readout electronics permits to reduce the 144 photosensor elements information to only 8+8 signals without significant detector performance degradation.

[0135] This approach also allows for resolving radioactive sources in the whole volume of the proposed crystal, but with a significant reduction of readout signals to be processed (from 144 to 16).

[0136] FIG. 9 depicts the results of an experiment carried out comparing the third alternative (projection reduction readout electronics) on a 1212 photosensor with a more used projection readout for both 1212 and 88 SiPM array. That means using 8+8, 12+12 and 8+8 signals respectively. One of the columns of sources was placed at 2.5 mm from the crystal edge. Both the projection reduction readout and the projection readout for the 1212 array made it possible to resolve these impacts at 2.5 mm from the edge. The 88 array with also 8+8 signals did not allow one for this. Therefore, the current invention shows how to reduce the number of channel without performance decrease at the edges,

[0137] FIG. 10 compares the two described embodiments, 1) and 2) providing the same result, as expected. We have plotted the measured spatial resolution as a function of the DOI layer (DOI layer width is 5 mm, where DOI1 is at the entrance). We observe comparable results for both approaches.