METHOD FOR CALIBRATING A IONIZING RADIATION DETECTOR AND DEVICE THEREFOR

专利摘要:
An object of the invention is a method of calibrating an ionizing radiation detector, for determining a correction factor to establish an amplitude-energy match. A first object of the invention is a method of calibrating a device for detecting an ionizing radiation, the detector comprising a semiconductor or scintillator detector material, capable of generating a signal S of amplitude A during a interaction of ionizing radiation in said detector material, the method comprising determining a weighting factor at amplitude A.
公开号:FR3030778A1
申请号:FR1463093
申请日:2014-12-22
公开日:2016-06-24
发明作者:Sylvain Stanchina；Guillaume Montemont
申请人:Commissariat a lEnergie Atomique CEA；Commissariat a lEnergie Atomique et aux Energies Alternatives CEA；
IPC主号:

专利说明:

[0001] TECHNICAL FIELD The invention relates to detectors for ionizing radiation, in particular for X-ray or gamma photon radiation, for applications in the nuclear, medical or non-destructive testing field. . STATE OF THE PRIOR ART Certain ionizing radiation detection devices use semiconductor detector materials of CdTe, CdZnTe type polarized by electrodes. Generally, a cathode and an anode are disposed on two opposite sides of the detector material. These electrodes allow the polarization of the detector material. They also allow the collection of charge carriers generated by the interactions of ionizing radiation in the detector material. Each interaction, whether it is a photoelectric interaction or a Compton scattering, or even a pair creation, generates several hundred or even thousands of electron-hole pairs. The electrons migrate to the anode (s) while the holes migrate to the cathode (s). Thus, an interaction is an event generating charge carriers in the semiconductor material, these charges migrating towards the electrodes polarizing the detector. Due to the higher mobility properties in the semiconductor materials currently used, the electrons, collected by the anode, produce an electronic signal, making it possible to estimate the energy lost by the ionizing radiation during the interaction, or energy of the event. Electronic circuits connected to the anode make it possible to amplify and shape the signal, in the form of a pulse, whose height or integral depends, often linearly, on the energy of the event. Each electrode, whether it is the cathode or the anode, can be segmented into several elementary electrodes. For applications related to imaging, the anode is generally divided into an elementary anode matrix, each elementary anode constituting a pixel of the detector. A difficulty related to this type of detector is that the signal collected by an electrode depends on the location of the event in the detector. For example, an interaction releasing the same energy, therefore the same number of charge carriers, does not produce the same signal if it occurs near the cathode near the anode. Moreover, in the case where an electrode is segmented into different elementary electrodes, the signal produced by an interaction depends on the position of the interaction with respect to the elementary electrodes. For example, if the elementary electrodes are distributed in a matrix, the signal produced by an interaction depends on the location of the interaction in the plane of the matrix. Thus, a signal collected by an electrode depends not only on the energy released by the interaction that generated this signal, but also on the position of the interaction in the detector material. In other words, the response of the detector is not spatially homogeneous and depends on the place of interaction.
[0002] The invention described in this application addresses this problem. SUMMARY OF THE INVENTION A first object of the invention is a method of calibrating an ionizing radiation detector, aimed at determining a correction factor as a function of the interaction location in the detector.
[0003] More specifically, a first object of the invention is a method of calibrating a device for detecting an ionizing radiation, the detector comprising a semiconductor detector material, extending between two electrodes, at least one electrode being able to generate a signal S of amplitude A during an interaction of ionizing radiation in said detector material, the method comprising the following steps: a) irradiating the detector using a known reference source, so as to generating interactions in the detector, b) selecting a plurality of correction coefficients j, and determining a plurality of reference spectra Spref ,, each reference spectrum being associated with a correction coefficient j, each reference spectrum representing a distribution of the amplitude A of the signal S generated when the detector is irradiated by said reference source Sref, the amplitude A of the signal S being weighted by said correction coefficient j, c) detecting a plurality of interactions in the detector during said irradiation and, for each interaction, measuring the amplitude A of an electronic signal S generated by at least one electrode and determining the value k of at least one position parameter of the interaction as a function of said signal S. d) for a plurality of values k of said position parameter, determining a calibration function Fk such that: - each calibration function Fk is associated with one of said k values of said position parameter, - each calibration function Fk is defined on a plurality of correction coefficients j, - each term Fk (j) of said calibration function represents a probability associated with said calibration function Fk correction coefficient j when the position parameter takes the value k, - each term Fk (j) being updated, for each interaction whose position parameter is equal to the value k at which said calibration function Fk is associated, by combining said amplitude A measured during this interaction with each reference spectrum Spred, e) determining a correction factor jk associated with each value k of the parameter of By identifying a remarkable point of said calibration function Fk, a correction function J associated with said position parameter has been established, with, for each value k of this position parameter, J (k) = jk. The calibration method then consists in associating, for a plurality of values k of said position parameter, a correction factor jk, each factor being associated with a value k. This factor is intended to be used during the operating phase of the detection device, when, under being exposed to ionizing radiation, the detector generates signals under the effect of interactions occurring in the detector material. Each interaction is assigned a value k of a position parameter, and the amplitude of the signal it generates is then weighted by the correction factor jk corresponding to said value k. This calibration can be performed in an automated manner, without recourse to a user, and this when the detector comprises a measurement channel or a multitude of measurement channels. It can also address simultaneously several position parameters, in particular a position of the interaction between the anode and the cathode and a position in the plane of the anode. It then takes into account the 3D position of the interaction. By correction function associated with a position parameter, is meant a function J whose variable k is the value of said position parameter, the correction function J (k) indicating, for each value k of said position parameter, the factor of correction jk to apply. The value k of the position parameter can be between a minimum value kmin and a maximum value kmax, such that kmin k kmax. kmin and kmax depend on the position parameter considered and can be predetermined.
[0004] The correction function can be set to a few discrete k values, and then interpolated between these discrete values, so that it is set to an interval [kmin - kmax]. In other words, the correction factors jk can be determined for a few discrete k values, between the minimum value k min and the maximum value kmax, the correction factor associated with a value lying between two discrete values then being obtained by interpolation, for example by linear interpolation. Step b) may comprise: the loading of an initial reference spectrum Spref, previously stored in a memory, corresponding to said reference source Sref, this Spref spectrum representing the distribution of the amplitude of the signal detected when the detector is irradiated by said reference source Sref, ii from the initial reference spectrum Spref, determining a plurality of reference spectra Sprefl, each reference spectrum Spref being obtained from the initial reference spectrum Spref, each value Spref, () of the reference spectrum being obtained from the value Spref (i / j). The term correction coefficient denotes a strictly positive number j, intended to modulate the reference spectrum Spref corresponding to the reference source, by normalizing the variable i, corresponding to the amplitude (or the energy) so as to establish a reference spectrum associated with the correction coefficient j Spref, such that Spref, (i / j) = Spref (i) × Knorm (j), the term Knorm (j) being a normalization term that can be equal to 1 or equal to 1 / d.
[0005] The correction coefficients j considered can be arbitrarily determined between a minimum value jrn ,,, and a maximum value Max.j These are discrete values, strictly positive, distributed between the values j and Max.j Preferably, jr. <1 <jmak. Step c) may comprise: at each interaction detected, determination of the amplitude A, A ', of at least two signals S, S' respectively generated by an electrode E, E 'and determination of the value k d a position parameter of the interaction in the detector as a function of these signals S, S ', and especially by combining the amplitudes A, A', for example in the form of a ratio A '/ A.
[0006] Step d) may comprise the following steps: initialization of each calibration function Fk associated with each value k of said position parameter, ii with each interaction detected, determination of the value k of said position parameter and update of the function corresponding calibration Fk, by adding, at each term Fk (j) of this function, the value of the corrected reference spectrum Spref; (A) at said amplitude A, iii repeating step ii until a predetermined stopping criterion is reached. The initialization of the calibration function Fk may especially be such that Fk (j) = Fkc) (j), each value Fa) representing the initial value of the term Fk (j). The set of initial values Fk ° (j) may be equal to the same number, for example 0. The predefined stopping criterion may be a predetermined update number for each calibration function Fk, or a duration of irradiation of the predefined detector, or a predefined number of detected interactions, or a maximum value reached by one or more terms Fk (j) of one or more calibration functions Fk In step e), the correction factor jk associated with the value k of the position parameter can be determined by identifying the value of j maximizing the value of the calibration function Fk, in which case: jk = argmaxi [Fk (j)] The method implemented according to the invention may include any of the features described below, taken singly or in any technically feasible combination. Among the electrodes connected to the detector material, there is at least one cathode and at least one anode. According to certain embodiments, the anode comprises a plurality of elementary anodes, in particular coplanar anodes, in particular in a matrix form, and / or the cathode comprises a plurality of elementary cathodes, in particular coplanar cathodes, in particular in matrix form. The position parameter of an interaction can be determined by combining signals resulting from said interaction, originating from different electrodes, and for example by an anode and a cathode, or by two elementary anodes, the two elementary anodes being in particular arranged in the same plane, or by two elementary cathodes, the two cathodes being in particular arranged in the same plane. The position parameter of an interaction can be determined by combining the signals of: the elementary anode having collected the maximum amplitude signal, called the winning anode, and the adjacent elementary anode of the winning anode having collected the signal of maximum amplitude, called secondary anode, in which case the value of the position parameter depends on the position of the interaction in the plane of the winning anode and the secondary anode. The position parameter of an interaction can in particular be obtained by making a ratio of the amplitude of the signal measured by the secondary anode to the amplitude of the signal measured by the winning anode.
[0007] The position parameter of an interaction can be determined by combining the signals: of the elementary anode which has collected the maximum amplitude signal, called the winning anode, and of the cathode or if the cathode is divided into different elementary cathodes, the elementary cathode having collected the maximum amplitude signal, in which case the value of the position parameter depends on the position of the interaction between the winning anode and the cathode, referred to as the depth of detection and frequently referred to by the acronym DOI ( Depth of Interaction). The position parameter of an interaction can in particular be obtained by making a ratio of the amplitude of the signal measured by the cathode to the amplitude of the signal measured by the winning anode. The initial reference spectrum Spref or each reference spectrum Spref preferably comprises one or a plurality of significant peaks isolated from each other. Preferably, the number of significant peaks Pi ... Pn constituting the reference spectrum is between 1 and 10. By significant peak is meant a peak whose height is greater than 2 times, preferably 5 times, the level of the background noise, on both sides of this peak. According to an example, step d) comprises an iterative update of the calibration function Fk such that for any j between kiwi and jmax, FkP (j) = Fk13-1 (j) + Sprefi (A) where: - Fk (j) is the value of the term Fk (j) preceding the iteration bead, - FkP (j) is the value of the term Fk (j) after the iteration bead, - A is the amplitude of the signal S measured by a electrode following an interaction whose position parameter is equal to value k. In this example, we consider the signal SAO measured by the winning anode AO, so A = AA0. p is the iteration index corresponding to the rank of the update. During the first update, p = 1. Step d) can then be repeated for a predetermined number of iterations, or until between two successive updates, the correction function Fk (j ) associated with the value k no longer significantly changes, or reaches, for a correction coefficient j, a predetermined value. In general, step d) is repeated until a predetermined stopping criterion is reached. According to a variant: the steps c), d) and e) are repeated iteratively for the same value k of the position parameter, each step e) resulting in the determination of a partial correction factor jkg, q designating the rank of the iteration, - and, following the first iteration, step d) comprises updating each term Fki (j) - (/ by weighting said amplitude A by the product of the partial correction factors jk jk , associated with the value k of the position parameter of the interaction, these factors having been determined during the previous q iterations, the method then comprises, during step e) of the last iteration, the determination of the correction factor jk associated with each value k of the position parameter as a function of the product of the partial correction factors determined by the successive iterations. Thus, a first determination of correction factors jk associated with a plurality of values k of a position parameter is carried out, by implementing steps a) to e) previously described, these partial correction factors being notedj. Then, the steps c), d) and e) are repeated, and, during the update of the calibration function Fk associated with each value k of the position parameter, the amplitude A of the signal S is weighted by the partial correction factor corresponding to said value k, previously determined. In general, during each iteration q, steps c), d) and e) are repeated, so as to obtain a partial correction factor jki and, during the next iteration, update the calibration function Fk associated with each value k of the position parameter is performed by weighting - (/ the amplitude A of the signal S by the products of the partial correction factors jk ... jk corresponding to said value k, previously determined. one embodiment, each interaction can be characterized by a plurality of position parameters, each position parameter of rank i having the value k, in which case the method implements steps a) and b) are analogous to those described above.
[0008] Step c) comprises determining a set of N position parameters of the interaction as a function of the signal S, N being an integer greater than 1, in which case step d) comprises selecting a parameter of position i among the N position parameters, and, for a plurality of values k, of said i 'position parameter, the determination of a calibration function Fki such that - each calibration function Fki is associated with one of said values ki of said position parameter, - each term Fki (j) of said calibration function represents a probability associated with said correction coefficient j when the position parameter takes the value ki, - each term Fki (j) being updated, at each interaction whose ith position parameter is equal to the value ki at which said calibration function Fki is associated, by weighting the amplitude A, measured during this interaction, by at least one correction factor j, led it is a correction factor j, being determined as a function of the value kr, of a ni 'position parameter characterizing the interaction, and combining the amplitude A thus weighted with each Spred reference spectrum.
[0009] By th and n 'position parameter is meant the position parameters of respective ranks i and n in the set constituted by the N position parameters characterizing each interaction. According to this embodiment, step e) comprises determining the correction factor associated with each value k, of the ith position parameter as a function of said calibration function Fki associated with the ith position parameter, and in particular according to a remarkable point of this function, for example its maximum. The correction factors jkr, are either previously determined or initialized to an initial value jkr, °. For example, during step e), the correction factor jk, associated with the value k, of the ith position parameter is obtained by the relation Ia, = argmaxi [Fia, (j)].
[0010] Thus, a correction function Ji associated with said ith position parameter has been established, with, for each value k, of this position parameter, .1, (1q) = iki- By correction function associated with an ith position parameter, is meant a function J whose variable k, is the value of said ith position parameter, the correction function .1, (1 (,) indicating, for each value ki, the position parameter According to this embodiment, the step d) may comprise: initialization of each calibration function Fki associated with each value k, of the ith position parameter ki, ii with each detected interaction: determination of the values (ki ... k, ... kN) of the N position parameters characterizing the interaction, weighting of the value of the amplitude A measured during this interaction by at least one correction factor jkr, depending on the value kr, of a characterizing position parameter interaction, addition, at each term Fki ( j) of said calibration function, the value of the corrected reference spectrum Spref, (A *), where A * represents the value of the weighted amplitude A, iii repetition of step ii until the achievement of a stop criterion. For example, during this step d), each function Fki, associated with the value ki of an ith position parameter, is updated iteratively according to the following expression: For all j between jrnin and imax, F / (J) = Ffd 1 (i) Spre fj (fl1 j kn XA) where: - FJ-1 (j) is the value of the term Fki (j) preceding the update, - FkiP (j) is the value of the term Fki (j) after the update, - A is the amplitude of the signal S measured by an electrode following each interaction whose I 'position parameter has the value k ,. In this example, the signal SAO measured by the winning anode AO is considered, so A = AA0, -k correspond to the correction factors corresponding to one or more position parameters characterizing the interaction, these parameters being either determined in advance or initialized to an initial value, as explained below, - p is the iteration index. At the first update, p = 1. According to this embodiment, when a correction parameter jkr, has not been previously determined, its value is assigned to an initial value, for example equal to 1. Thus , prior to the steps described above, the method comprises an initialization phase in which each correction parameter jkr is assigned an initial arbitrary value, for example equal to 1. In one embodiment, the steps c ), d) and e) previously described, taking into account a plurality of position parameters, are carried out successively by selecting a different position parameter each time. In other words, these three steps c), d) and e) are carried out: for a first position parameter so as to establish a correction function Ji associated with said first position parameter, with, for each value k1 of this parameter of position, Ji (ki) = jki, step d) being implemented by considering the correction factors jki ... jkN equal to an arbitrary initial value jki ... jkN °, for example equal to 1, - for an nth position parameter, with 1 <nl l, so as to establish a correction function Jr, associated with said nth position parameter, with, for each value kr, of this position parameter, .1 ,, (kr, ) = jkn, step d) being implemented by weighting the amplitude A measured by correction parameters jki ... jkr, _i previously determined, and correction parameters equal to an arbitrary initial value, for example 1 .
[0011] According to one variant: steps c), d) and e) are repeated iteratively, for the same value ki of the ith position parameter each step e) resulting in the determination of a factor of qi partial correction jki, qi denoting the rank of the iteration, and, following the first iteration, step d) comprises updating each term Fki (j) by weighting said amplitude A by the product of the partial correction factorsji ... jk, associated with the value ki of the ith position parameter of the interaction, these factors j1 qi ki ... jki having determined over the ci; previous iterations. The method then comprises, during step e) of the last iteration, the determination of the correction factor jki associated with each value ki of the ith position parameter as a function of the product of the partial correction factors jki jki determined during the O .; successive iterations. Q denoting the number of iterations required. In a general manner, according to this variant, during each iteration qi, steps c), d) and e) are repeated so as to obtain a partial correction factor jkqii and, during the next iteration, the setting update of the calibration function Fki associated with said value Here of the ith position parameter is performed by weighting the amplitude A of the signal S by the product of the partial correction factors .qi jki ... jki associated with the value ki of the I 'position parameter, previously determined. The product ... j corresponds to an intermediate correction factor associated with the value ki of the ith position parameter. Preferably, during step d), the calibration function Fki is updated by also weighting the amplitude A by a weighting factor K, this factor being obtained from the product of the correction factors j "or correction factors jilin ... jkg: partial, associated with kr values, other position parameters of the interaction. A second object of the invention is a device for detecting an ionizing radiation, enabling the implementation of the previously described method. Such a device is understood in particular: a) a detector, comprising a semiconductor detector material, extending between two electrodes, the detector being able to generate charge carriers under the effect of an interaction with the ionizing radiation, b) an electronic processing circuit, able to form a signal S from charge carriers collected by an electrode during an interaction and to determine the amplitude A, c) an electronic location circuit, able to determine the value of at least one position parameter at said interaction in the detector material from different electrodes, d) an electronic calibration circuit, configured to implement steps b), c) and d) and e) of method described above. A third object of the invention is a method for detecting an ionizing radiation, using a detector, in particular a semiconductor detector extending between two electrodes, the method comprising the following steps: a) irradiation of a detector by means of an ionizing radiation source, b) at each detection of an interaction of the ionizing radiation in the detector, formation of at least one signal S, under the effect of said interaction, the signal S being collected by an electrode E, and determining the values ki ... kN of N parameter (s) of the position of the interaction, N 1, from said signal S, c) determination of at least one factor of correction jk, as a function of the value ki ... kN of each correction parameter, d) correction of the amplitude A of a detected signal S by applying to it at least one correction factor jk, the latter being determined according to of the value k, of each position parameter characterizing the intera ction. Each correction factor jk associated with a given position parameter is then preferably established according to the previously described calibration method.
[0012] The correction of the amplitude is in particular carried out by a weighting of said amplitude by the each correction factor jk, corresponding to the value k, of the I 'position parameter. Preferably, the set of correction factors, associated with each position parameter, are previously stored in a table, called a correction table, defining, for each correction parameter, a correspondence between the value k, of this parameter and the correction factor jk, corresponding to this value. According to a variant, step b) comprises the determination, at each detected interaction, of the value of N position parameters with N 2, step c) comprises the correction of the amplitude of the detected signal S by applying to it a plurality of correction factors jki .... jkN, each correction factor being determined as a function of the respective values k1 ... kN of the position parameters, According to one embodiment, each correction factor jk, is a multiplicative term applied to the amplitude A of the signal produced by the interaction. In other words, step c) comprises multiplying the amplitude A of the signal detected by one or more correction factors jk ,, each correction factor jk, depending on the value k, of the position parameter considered. We then obtain a corrected amplitude A * such that A * = x A. For example, the position parameters of an interaction are respectively a coordinate (XINT, YINT) in the plane of the anode and a coordinate (ZINT) between the anode and the cathode, as previously mentioned. In this case, the amplitude of each detected signal is corrected according to the correction factors jki and jk2 each correction factor jki and jk2 depending on the respective values 1 (1, k2 of each position parameter considered. corrected amplitude A * is such that A * = jkl X jk2 X A. A fourth object of the invention is a device for detecting an ionizing radiation comprising: a) a detector, comprising a semiconductor detector material, s' extending between two electrodes E, the detector being able to generate charge carriers under the effect of an interaction with the ionizing radiation, b) an electronic processing circuit capable of forming a signal S from charge carriers collected by an electrode E during an interaction and to determine its amplitude A, c) an electronic location circuit, able to determine the value k of at least one position parameter of the interaction in the material to the detector from at least one signal collected by an electrode E, d) an electronic correction circuit configured to apply, to a detected signal S, a correction factor jk corresponding to said value k of said position parameter, said factor correction being previously stored in a memory.
[0013] The correction factor jk can be determined by the calibration method previously described. Preferably, the correction electronic circuit applies the correction factor to a single detected signal S, for example to the anode, or, when the anode has a plurality of elementary anodes, to the winning anode. According to one embodiment, step c) comprises determining the value of at least N position parameters, N being an integer greater than 1, for example the interaction depth ZINT as well as the position of the interaction (XINT, VINT) in the plane of the anode. In this case, during step d), the detected signal S is corrected by N correction factors ji jN, each correction factor depending on the value of the position parameter with which it is associated. In particular, the amplitude of the detected signal is weighted by a product of the N correction factors.
[0014] DESCRIPTION OF THE FIGURES FIG. 1 represents a diagram of the detection device; FIG. 2A represents the different steps of the calibration method according to a first embodiment of the invention; FIG. 2B represents the different steps of the calibration method according to a variant of a first embodiment of the invention; Fig. 3 shows an example of Spref reference spectrum; Figure 4 shows a plurality of corrected reference spectra Spref,; Fig. 5 is a 3-dimensional view of a plurality of Spref-corrected reference spectra; ; FIG. 6 represents a calibration function Fk, when the position parameter k corresponds to the position of the interaction between a cathode and an anode; FIG. 7 represents an example of the evolution of the correction factor jk as a function of the values k of the position parameter, when the latter corresponds to the position of the interaction between a cathode and an anode; FIG. 8 represents an example of the evolution of the correction factor jk as a function of the values k of the position parameter k, when the latter corresponds to the position of the interaction in the plane of the anode, the latter being segmented into elementary anodes; FIG. 9 represents the different steps of the calibration method according to a second embodiment of the invention; FIG. 10 represents the different steps of the calibration method according to a third embodiment of the invention. DETAILED DESCRIPTION OF PARTICULAR EMBODIMENTS FIG. 1 represents a device 1 for detecting ionizing radiation according to the invention. It comprises a detector 10, made of a semiconductor material 11, capable of being polarized by a cathode 12 and a matrix of elementary anodes 13. This anode matrix consists of elementary anodes 131 ... 13p. In this example, the semiconductor material is CdTe. It may be another semiconductor material commonly used for the detection of ionizing radiation (Si, CdZnTe ...). The thickness of the material is a few millimeters, for example 5 mm. In a CdTe type material, the electrons have more favorable charge mobility properties. Also, the signals giving the most accurate information on the energy deposited by an interaction in the detector are those coming from or elementary anodes 131 13k ... 13p.
[0015] In this example, the cathode 12 is unique, but as for the anode, a division into multiple elementary cathodes would be possible. Each elementary anode 13k is made by disposing a conductive material on the surface of the semiconductor material. The size of each elementary anode 13x is for example (to be completed), the spacing between each elementary anode being 50 nm to 3 mm, preferably between 100 nm and 1 mm. In a symmetrical manner, the cathode 12 is made by disposing a conductive material on the surface of the semiconductor material, generally the surface opposite to the surface hosting the anode.
[0016] In this example, the detector material 11 is a CdZnTe parallelepiped of thickness 5 mm and 20 mm of sides. The device also comprises: a first amplification electronic circuit 23x associated with each elementary anode 13x so as to produce an anode signal whose amplitude is proportional to the number of charge carriers collected by the elementary anode as a result of of an interaction. Such a circuit is known and will not be described in detail, a second electronic amplification circuit 22, associated with the cathode 12 so as to produce a cathode signal whose amplitude is proportional to the number of charge carriers collected by the cathode as a result of an interaction. Such a circuit is known and will not be described in detail. Such a circuit is optional. Thus, in general, each interaction in the detector causes the collection of charge carriers by at least one electrode E, (anode A or cathode K) so as to generate an electronic signal S whose amplitude is A.
[0017] The term amplitude refers to the maximum height of a signal when the latter is in the form of a pulse. It can also be the integral of this signal, or any other function of the maximum height or the integral. In general, the amplitude represents the amount of charges collected by the electrode E during an interaction. An interaction frequently leads to the collection of charge carriers by several electrodes E, E ', E "... The amplification circuit associated with each of these electrodes then producing signals S, S', S". respective amplitudes A, A ', A "The term" winning "electrode designates the electrode E producing the signal S of maximum amplitude A. Each amplification circuit (23x, 22) may comprise an amplifier , affecting the electronic signal of a gain G. The gain of each amplification circuit is determined beforehand by those skilled in the art.
[0018] The detection device 1 also comprises an electronic positioning circuit 30, able to determine the value k of a position parameter of the interaction in the detector 11. By position parameter, is meant a parameter, determined using at least one signal S detected an electrode, and depending on the position of the interaction in the detector 11.
[0019] This position can be: a position between the cathode 12 and the anode 13, that is to say a ZINT coordinate of the interaction along an axis extending between the cathode and the anode. This position, called the depth of interaction, and whose acronym is DOI (Depth of Interaction), translates the distance ZINT between the interaction and the anode (or the cathode), ii a position in the plane of the anode 13, that is to say a pair of coordinates (XINT, YINT) of the interaction in the plane of the anode. The electronic positioning circuit 30 determines the value k of the position parameter of an interaction by using the electronic signals S, S ', S "generated by one or more electrodes E, E', E". The value of this position parameter is generally calculated using at least the signal S of the affected electrode, in particular the affected anode, but this is not imperative. The position parameter k can be defined using a single electronic signal S: for example, if a single anode AO, among the different elementary anodes 13x, collects a significant signal S, the value k of the position parameter corresponds to at the position of the anode AO. It is then considered that the position of the interaction in the plane of the anode 13 corresponds to the coordinates (X0, Yo) of the center of the anode A0, these coordinates (X0, Yo) constituting the value of the position parameter. By significant signal is meant a signal whose amplitude is greater than a certain threshold, the latter being determined as a function of the electronic noise, so as to avoid false detections.
[0020] However, when the anodes are sufficiently close to each other, the charge carriers can be distributed over several adjacent 13x elementary anodes. The elementary anode collecting the signal whose amplitude is maximal is designated by the term "winning anode" (or anode touched), and denoted AO. As in the previous paragraph, the value of the parameter the position of the interaction in the plane of the anode corresponds to (X0, Yo).
[0021] The value k of a position parameter can be determined by combining at least two signals S, S 'respectively detected by different electrodes E, E'. It can be a ratio between the amplitude of the signal of the cathode K, on the amplitude of the signal of the affected anode AO. The position parameter then indicates a position of the interaction between the winning anode AO and the cathode K, which corresponds to the previously defined ZINT interaction depth. It can be a ratio between the amplitude of the signals between two adjacent anodes, and in particular the signal of the secondary anode A1 and the signal of the winning anode AO, the secondary anode corresponding to the anode, adjacent to the winning anode, whose signal has the highest amplitude.
[0022] The lower the AAJAAo ratio, the closer the position of the interaction, in the plane of the anode 12, is to the center of the winning anode AO. The higher this ratio, the more the interaction position moves away from the center of the ANI anode towards the center of the anode Am. When the ratio is equal to 1, the interaction is located between the anode winner and the secondary anode. When this ratio is 0, the interaction is centered relative to the winning anode.
[0023] Naturally, several signals S, S ', S ", S- derived from adjacent electrodes E, E', E", E "can be combined to obtain an even more precise determination of the position of the interaction in the plane of The detection device 1 also comprises an electronic calibration circuit 40 capable of implementing the detector calibration method, which circuit is an important element of the invention and will be described hereinafter. Detection 1 comprises a memory 50, comprising the correction factors determined by the calibration method detailed below, For each correction parameter, a correction factor j k is associated for different values k of said correction parameter. also comprises an electronic correction circuit 60, able to apply, to an electrode signal S, a correction factor jk depending on the value k of at least one parameter of positio n of each interaction in the semiconductor material 11. These factors can in particular be derived from the memory 50. In general, a correction factor is assigned to the amplitude A of the signal S of the winning anode AO (or, the where appropriate, at the signal of the winning cathode KO). This amplitude is then corrected according to the value k of said position parameter, as explained later.
[0024] The detection device is able to detect ionizing radiation. The term ionizing radiation denotes radiation capable of generating charge carriers in the detector 11. It may in particular be X, y, charged particles (in particular a or p) or neutrons.
[0025] The amplification circuits 22, 23 and the location circuit 30 are known and will not be described in detail. The calibration method, implemented by the electronic calibration circuit 40, will now be described, with reference to FIG. 2A. Its function is to establish, for different values k of a position parameter, a correction factor jk, this correction factor being intended to be applied, subsequently, to the amplitude of a signal S collected by a electrode, and more particularly by the winning anode, when this signal results from an interaction whose value of the position parameter is equal to k. Thus, an objective of the invention is to allow the weighting of the amplitude of a detected signal by a correction factor determined as a function of the position of the interaction in the detector, in order to take into account the heterogeneity spatial response of the detector. The electronic calibration circuit 40 is configured to perform a calibration of the detection device 1, so as to correct the spatial heterogeneity of its response. This calibration circuit is implemented when the detection device 1 is in a calibration mode. The basic idea of the calibration is to irradiate the detection device 1 with the aid of a source of known ionizing radiation, referred to as the reference source Sref, for which the theoretical response of the detection device is known. It is then a question of analyzing the signals produced in response to this irradiation, and of determining an appropriate correction, on the basis of the theoretical response of the detector towards this reference source, this theoretical response being qualified as a reference response. .
[0026] FIG. 2 represents the steps of a first embodiment of the calibration method. In a first step 100, a reference source Sref is placed in front of the detection device 1, knowing that the response Sp'f of the device, in this configuration, is known. In this example, the detector response is an energy spectrum produced by the detector in response to the source. By energy spectrum is meant the amplitude distribution of the signals S from an electrode E during the exposure of the detector to a radiation source. Such a spectrum is in the form of a histogram Sp, each term Sp (i) represents the number of signals detected by the electrode in question, and whose amplitude is equal to i. By amplitude equal to i; is meant an amplitude within a range of amplitude Ai including the value i, and for example centered around the value i.
[0027] This response can be determined by modeling the detector. It can also be determined experimentally, by adopting an acquisition time long enough to be statistically representative. The responses obtained on each elementary anode 13x can then be averaged to form a reference response Spref of the detector facing the source. Alternatively, the response can be determined by selecting the interactions having a determined position parameter, and realizing the amplitude spectrum of the signals collected during these interactions. In the example considered, the reference response Spref is an energy spectrum corresponding to the spectrum "theoretically" produced by the detector 11 in response to the reference source Sref. FIG. 3 represents such a Spref reference spectrum. It comprises several peaks P1, P2, P3, P4 which are different from the background noise B. It is preferable that the initial reference spectrum Spref comprises at least one peak, and preferably between 1 and 20 peaks. Preferably, the width at half height of a peak is less than one-tenth, or even twentieth or thirtieth of the dynamics of the spectrum, the latter corresponding to the amplitude range in which the spectrum is recorded. Preferably, each peak has a height greater than at least two times, better at least five times the level of the background noise on either side of the peak. The example in Figure 3 represents the spectrum of a source of 'Co. We distinguish the photoelectric peak, which corresponds to the emission of 122 keV gamma photons, as well as the X fluorescence peaks of the material constituting the collimator of the source, in this case Tungsten. Such a spectrum Spref can be likened to a probability distribution of measuring a given amplitude knowing the source radiating the detector. The first step 100 of the calibration process is to load a reference spectrum Spref, corresponding to a reference source Sref. This reference spectrum is referred to as the initial reference spectrum. The second step 200 aims to establish, from the initial reference spectrum Spref, a plurality of Sprefl reference spectra, each associated with a correction factor j. Each spectrum Spref corresponds to the initial reference spectrum Spref by taking into account a correction factor equal to j. A correction coefficient j is a strictly positive coefficient, between a minimum value jmir, and a maximum value jmax. kiwi and jmax are arbitrarily defined. In this example, 0.82. Preferably, the interval jrnin <1 and jmax> 1. The values of j are selected are predetermined.
[0028] More precisely, if Sore (i) designates the value of the spectrum Sore corresponding to the amplitude channel i, the reference spectrum associated with the correction coefficient j can be written Spref, (i) = Sore (i / j). Each, each channel of a spectrum Sore, represents the number of collected signals whose amplitude is situated in an interval -Ai centered around the value -. The previous equality may also include a Knorm normalization term (j), which may be equal to 1 or equal to 1 / d, in which case] (i) = Spref (in). Such normalization makes it possible to preserve the integral of the Spref spectrum. When the correction coefficient j is greater than 1, the reference spectrum Spref corresponds to a compression of the initial reference spectrum Spref, insofar as the information of the spectrum is concentrated on a number of channels smaller than the number of channels constituting the Spref spectrum. Conversely, when the correction coefficient j is less than 1, the reference spectrum Spref corresponds to a dilation of the initial reference spectrum Spref, the information of the spectrum being dilated according to a number of channels greater than the number of channels constituting the spectrum. initial reference Spref.
[0029] FIG. 4 represents several Spref reference spectra, associated with a correction coefficient j between 0.8 and 2. It is specified that in FIG. 4, the abscissa scale is not linear. Each column of FIG. 4, corresponding to an abscissa j, represents a reference spectrum Spref. Thus, at each point (j, i) is assigned a weight 131, indicated by the gray level scale, this weight representing the value of the spectrum Sore, at the amplitude i, so that Pli = Sprefj ( i). Thus, each column of FIG. 4 represents a probability distribution of the amplitude measured by the detector, knowing the correction coefficient j, when the detector is irradiated by the reference source Sref, the correction coefficient j being the abscissa. of the column. Moreover, each line of FIG. 4, corresponding to an amplitude i, represents a probability distribution of the correction coefficient j, when this amplitude is measured, the detector being irradiated by the reference source Sref. In other words, the detector being irradiated by the reference source Sref, if a signal S of amplitude A is measured, the distribution formed by the set of values Sprefj (A), for j between jmax and jmax, corresponds to a distribution the probability of the correction factor j to be applied, knowing the measurement of this amplitude A. By constructing a plurality of reference spectra Spref ,, each reference spectrum Spref, respectively corresponding to an abscissa column j of FIG. can establish a plurality of probability distributions, a correction coefficient j, knowing a measured amplitude A. These distributions will be used in the next step. FIG. 5 represents a 3-dimensional representation of FIG. 4. Each profile, corresponding to a fixed correction coefficient value, represents a Sprefp reference spectrum. The third step 300 consists in irradiating the detector using a radiation source of FIG. reference Sref corresponding to the Spref reference spectrum. During each interaction, an electrical signal, or a plurality of electrical signals (S, S ', S "), are collected across one or more electrodes (E, E', E"), and processed by the circuit amplification attached to each electrode. As previously described, the value k of a position parameter is assigned to each interaction detected as a function of the signal S or of the signals S, S ', S ".The position parameter can correspond to a position of the interaction in the plane of the anodes, in which case the value corresponds to a coordinate (XINT, VINT) of the interaction in the plane of the anode The position parameter can correspond to a ZINT interaction depth, in which case an interaction is close to the anode, the more it is said to be deep, in which case the value of the position parameter corresponds to a coordinate of the interaction along an axis perpendicular to the anode and the cathode. that the position parameter represents a position (XINT, YINT) in the plane of the anode 13 when the latter is divided into elementary anodes 131 ... 13p coplanars.When these elementary anodes are sufficiently close to each other, the Electrons generated by an interaction are collected by several 13x elementary anodes, generally adjacent. The collection of these signals is simultaneous, and each processing circuit 23 ', connected to an elementary anode 13' having collected a significant amount of charges, delivers a signal S of amplitude A. These signals S are addressed to the location circuit 30 which determines the anode AO having collected the maximum signal, called the winning anode, and the anode A1, called the secondary anode, whose signal SAi is the one whose amplitude Am is maximal among the different adjacent anodes of the winning anode. AO. The location circuit 30 determines a ratio between the amplitude of the signal delivered by the secondary anode, denoted by Am, and the amplitude of the signal delivered by the winning anode, denoted AA0. The ratio Acd / AA0 corresponds to a load sharing coefficient. It is possible to establish a function g, assigning a value k of the position parameter of the interaction according to this ratio. In other words, k = f (AA1 / AA0). More precisely, if the centers of the winning anode AO and of the secondary anode A1 respectively have coordinates (X0, Y0) and (X1, Y1), the coordinates (XINT, YINT) of the interaction INT are located on the line passing through (X0, Yo) and (Xi, Yi), the distance from the center (X0, Yo) of the winning anode AO depending on the position parameter k. The lower this is, the greater the amount of charge carrier collected by the winning anode AO compared to the secondary anode A1, plus (XINT, YINT) is close to (X0, Yo). Thus, the position of the interaction is determined: on the one hand by the coordinates of the center (X0, 110) of the winning anode A0 and the center (X1.1 (1) of the secondary anode A1, on the other hand by the value k of the ratio A1 / A0: for example, when A1 / A0 = 0, (XINT,, X0-FX1YO + Y1 VINT) = (X0Y0) and when A1 / A0 = 1, (X. --INT, YINT) = l 2 '2) "The location circuit 30 then calculates the value k of the position parameter of the interaction, by making a ratio between the amplitudes Am and AA0.
[0030] The fourth step 400 consists of establishing a calibration function Fk associated with different values k of the position parameter. More precisely, during this step, the calibration function Fk associated with the value k of the position parameter of the interaction detected in the previous step 300 is updated.
[0031] Initially, each calibration function Fk is initialized to a predetermined value, for example to a zero value: whatever j, Fk (j) = 0. This fourth step of the method consists in updating each function Fk so that whenever an interaction is detected, whose position parameter is equal to k, the value of the corrected reference spectrum Spref, corresponding to the amplitude A of the measured signal, is added to each term Fk (j). at each interaction, and that for j included between jrj and jmax. In other words, for each interaction detected, the value k of the position parameter of this interaction is determined, and then the calibration function Fk, associated with the position parameter k, is updated according to the iterative equation: For all j between jr. and jmax, Fk P (j) = Fk13-1 (j) + Spref, (A) where: - FkP-1 (j) is the value of the term Fk (j) preceding the iteration i, - FkP (j) is the value of the term Fk (j) after the iteration i, - A is the amplitude of the signal S measured by an electrode following an interaction whose position parameter is equal to k. In this example, we consider the signal SAO measured by the winning anode AO, so A = AA0. This iterative update is performed for a plurality of detected interactions, having a position parameter of the same value k, until a stop criterion is reached, for example a predetermined number of iterations or a value reached by one or more values of the function Fk. FIG. 6 represents a function Fk carried out after a number of updates equal to 1000, for a position parameter whose value k is equal to 0.35. The abscissa represents the correction coefficients j, between jmin and jmax. The ordinate axis represents the values of the calibration function Fk (j).
[0032] For different values k of this position parameter, the calibration circuit 40 establishes a calibration function, noted Fk, each function Fk representing the probability distribution of the correction factor j to be applied when the position parameter of an interaction is equal. at k. More precisely, each term Fk (j) is proportional to the probability that the correction factor, to be applied to each interaction whose position parameter is equal to k, is equal to j.
[0033] The fifth step 500 is the determination of the correction factor jk associated with each value k of the position parameter, as a function of a remarkable point of the calibration function Fk associated with this value k, and for example as a function of the correction coefficient j for which the function Fk takes a maximum value.
[0034] In this example, jk = argmax [Fk (j)]. Referring to the example of FIG. 6, this calibration function takes its maximum value for j = 1.08. Thus, according to this example, jk = 1.08. This means that when the position parameter has the value k = 0.35, the correction factor to be applied to the amplitude AA0 of the signal of the winning anode is jk = o, 35 = 1.08. FIG. 8 represents the correction function J associated with the correction parameter, each term J (k) representing the correction factor jk to be adopted as a function of the value k of the position parameter, the latter representing the position (XINT, VINT) , with respect to an elementary position anode (X0, 110) of the interaction in the plane of the anodes as previously explained. Each correction factor jk thus determined is stored in a memory 50, which is the subject of the sixth step 600. The storage can be performed in the form of a table, called a calibration table, making a correspondence between the value k of the position parameter and the correction factor jk. According to a variant of this embodiment, represented in FIG. 2B, the steps c) to e) can be carried out iteratively, for the same value k of the position parameter, the correction factor, corresponding to the value k of the position parameter, determined at each iteration of rank q being called partial correction factor and noted jkq. Beyond the first iteration, during the update of the calibration function Fk associated with each value k of the position parameter, the amplitude A of the signal S is weighted by the product of the correction factors jkl ... jkq corresponding to said value k, determined during the previous iteration or during the previous q iterations, q denoting the rank of the iteration. Such a product jklx ... x jkq corresponds to an intermediate correction factor associated with the value k. During step e) of the last iteration, q = Q, Q designating the number of iterations, the correction factor jk associated with the value k of the position parameter is calculated by producing the product of -Q correction factors partial jk .... jk successively acquired in each series. It can be seen that as iterations proceed, the partial correction factors gradually tend towards the value 1.
[0035] Thus, before the first iteration, the correction factor associated with the value k, denoted j2, is initialized to a predetermined value, for example equal to 1. Each iteration q results in the determination of a partial correction factor jkg, each partial correction factor comprising the factor of = 1 q 1, intermediate correction H / = 1 jk used as the weighting term of the amplitude measured during the next iteration, of rank q + 1. During step e) of the last iteration, it is possible to determine jk such that: rE1 = Q Jk-111 = 1.1k In other words, during step e) of the last iteration, the correction factor jk associated with the value k of the position parameter is determined as a function of the product of the partial correction factors corresponding to the same value k of the position parameter, these partial correction factors having been determined during the previous iterations. For example, if a series designates the determination of partial correction factors jkg by applying the steps c) to e) previously described, in each series, step d) comprises an iterative update of the calibration function Fk such that: for all j between in *, and imax, where: F713,41 (j) = F13, 141 (j) + Sprerial (riiik x A) - FkP-1 (j) is the value of the term Fk (j ) preceding the 13 'iteration, in the row series q (the term series here denotes the iteration q allowing the determination of the partial correction factor jig,), - Fkm (j) is the value of the term Fk (j) after the iteration bead, during the row series q, - A is the amplitude of the signal S measured by an electrode following an interaction whose position parameter is equal to k. In this example, consider the signal SAO measured by the winning anode AO, so A = AA0, - p is the iteration index corresponding to update of the calibration function. At the first update, p = 1, - jk is the partial correction factor associated with the value k of the position parameter, determined during series 1, with IS. Steps c) to e) are then renewed until to the achievement of a stopping criterion: it can be a predetermined number of series, a small difference between two partial correction factors jkg and .q + 1 jk successive.
[0036] Typically, according to this variant, the number Q of series is between 2 and 10. According to a second example, the position parameter represents the interaction depth ZINT, determined according to the signal of the winning anode and the signal of the cathode. More precisely, the location circuit 30 determines the winning anode by comparing the different signals emitted by the electronic circuits 23x. Then, it determines the ratio between the amplitude AK of the cathode signal, supplied by the cathode electronic circuit 22, on the amplitude AA0 of the signal of the winning anode. The locating circuit 30 then calculates the value k of the position parameter of the interaction, by effecting a ratio between the amplitudes AK and AA0, as described in the He, Zhong "1-D position sensitive single carrier semiconductor detectors", Nuclear Instruments and Methods in Physics Research, A 380 (1996) 228-231. It is possible to establish a function g, assigning a depth of interaction in the detector according to this ratio. In other words, k = g (AK / AA0). The position parameter can take different values k, between a minimum value kmin (adjacent interaction of the anode) and a maximum value kmax (adjacent interaction of the cathode). The calibration method according to this second example reproduces the same steps as that of the preceding example: step 100: loading of a reference spectrum Spref, corresponding to a reference source Sref. Second step 200: establishing a plurality of Sprefi-corrected reference spectra, these reference spectra being obtained from the Spref reference spectrum, by the Spref relationship; (i) = Spref (i / j). Preferably, a normalization term is introduced such that: Spref (i / j) Sprefj (i) = These first two steps are analogous to those of the first example. 3 'step 300: irradiating the detector using the reference source Sref, and determining, at each interaction, the value k of the position parameter and the amplitude A of the signal S measured by an electrode. In this example, the amplitude A is the amplitude AA0 of the signal SAO measured by the winning anode A0, as well as the amplitude AK of the signal AK measured by the cathode. 4th step 400: With each interaction detected, update of the calibration function Fk, where k is the value of the position parameter. The update of the function Fk is carried out according to the same principles as in the preceding example, and for example according to the iterative expression: For all j between kiwi and jmax, FkP (j) = Fk13-1 (j ) + Sprefi (A) where: - FkP-1 (j) is the value of the term Fk (j) preceding the iteration p, - Fk (j) is the value of the term Fk (j) after iteration p, -A is the amplitude of the signal S measured by an electrode following an interaction occurring at the coordinate k. In this example, we consider the signal SAO measured by the winning anode AO, so A = AA0. As in the previous example, each calibration function Fk is initialized to an arbitrary value, the set of values of this function can be initialized to 0. 5th step 500: determination of the correction factor jk associated with each value k the position parameter from the calibration function Fk. Similarly to the first example, each factor jk can be such that: jk = argmax [Fk (j)]. FIG. 7 represents, for different values k of the position parameter considered, the correction function J associated with this position parameter, each term J (k) of this function corresponding to the correction factor jk to be applied when the value of the parameter of considered position is equal to k. In this figure, the position parameters are normalized and lie between a value kmin = 0 and a value kmax = 1. 6th step 600: storage of the correction factors jk corresponding to each value k of the position parameter considered. According to a third example, represented in FIG. 9, two calibrations are successively carried out: i a first calibration, called depth calibration, by determining first correction factors jki as a function of the value k1 of a first correction parameter corresponding to the interaction depth, that is to say the distance of the interaction with respect to the anode 13, ii A second calibration, called calibration in the plane of the anode, by determining second correction factors jk2 as a function of the value k2 of a second correction parameter, depending on the position of the interaction in the plane of the anode 13. The advantage associated with this third example is that it has a pair of correction factors (jki, jk2) as a function of the position of the interaction respectively according to the depth of the detector and in the plane of the anode 13. The correction factors (jki, jk2) then take into account the position, in 3 dimension s, the interaction in the detector material 11. The order in which these calibrations are made does not matter: the first calibration, depending on the depth of interaction, can be performed before or after the second calibration, depending on the position in the plane of the anode 13. However, such a calibration can be further improved. Indeed, when two successive calibrations are performed according to two different position parameters (k1, k2), these two calibrations are performed independently of one another. During the deep calibration, the position parameter, defined from the K / AO ratio, is itself affected by an error, related to the position of the interaction in the plane of the anode 13, since it is determined at using the signal of the winning anode, the latter having not undergone any correction. Similarly, during calibration in the plane of the anode, defined from the ratio A1 / A0 is affected by the interaction depth, since it is determined using the signal of the winning anode, the latter has not been corrected.
[0037] In order to achieve a more accurate calibration, an iterative calibration is performed, as described in the fourth example. According to this fourth example, represented in FIG. 10, steps 100 and 200 previously described are carried out, then, alternatively, a first position parameter is considered, for example the interaction depth, defined by the ratio Ak / AA0. This makes it possible to establish a set of first correction factors jki, relative to the first correction parameter k1, ii a second position parameter, for example the position of the interaction in the plane of the anode, defined by the ratio AAJAAo. This makes it possible to constitute a set of second correction factor bases jk2, relative to the second correction parameter k2.
[0038] Then, iteratively, a step 300 'is carried out, considering said first position parameter, this step being similar to the step 300 previously defined. However, during step 300 ', at each interaction detected, the value k2 of the second position parameter is also determined. During the update of the calibration function Fu., Each measured amplitude is weighted by the correction factor jk2, associated with the value k2, the second correction parameter. The correction factors jki and jk2respectively associated with the set of values of the first and second correction factors are first initialized to an arbitrary value, for example 1.
[0039] More precisely, step 300 'comprises: irradiating the detector using the reference source Sref; at each interaction, determining the value 1 (1 of the first position parameter and the value k 2 of the second position parameter, and update of a calibration function Fkl associated with said value k1 of the first position parameter, according to the formula: For all j between kiki and imax, Fk (j) = Fk1134 (i) + Sprefj (jk2x A) where: - Fk113-1 (j) is the value of the term Fu (j) preceding the iteration p, - Fk (J) is the value of the term Fu (j) resulting from the iteration p, - A is the amplitude of the signal S measured by an electrode following the interaction In this example, we consider the signal SAO measured by the winning anode AO, so A = AA0, - ik2 is the factor correction value corresponding to the value k2 of the second position parameter, this parameter being either initialized to an arbitrary value jk20 (for example 1) or beforehand As in the previous examples, each calibration function Fk1 is initialized to an arbitrary value, the set of values of this function can be initialized to 0. The following step 400 'is similar to step 400 previously described. It includes updating the correction factors jki associated with the position parameter k1, as a function of the calibration function Fk1 associated with this position parameter k1: jki = argmax; [Fki (j)]. We then determine the correction function Ji, such that Ji (ki) = symmetrically, a step 300 "is carried out, considering said second position parameter, this step being similar to the step 300 previously defined. of step 300 ", the value k1 the first position parameter is also determined for each interaction detected. When updating the calibration function FK2, associated with the value k2 of the second parameter, the value k1 of the first correction factor jki, associated with the first correction parameter, is used as a weighting term. More precisely, step 300 "comprises: i irradiating the detector using the reference source Sref, ii At each interaction, determining the value k1 of the first position parameter and the value k2 of the second parameter of position, then update of a calibration function associated with the value k2 of the second position parameter, according to the formula: For all j between jrj and jmax, Fk2Pi (j) Fk2P W + Sprefj (jki XA) where : - Fk2P-1 (j) is the value of the term Fk2 (j) preceding the update, - Fk2 P (j) is the value of the term Fk2 (j) after the update, - A is the amplitude of the signal S measured by an electrode following the interaction In this example, the signal SAO measured by the winning anode A0 is considered, therefore A = AA0, - jki is the correction factor corresponding to the value k1 of the first parameter of position 1 (1, previously determined, as in the preceding examples, each calibration function Fk2 is initialized to an arbitrary value, the set of values of this function can be initialized to 0. The next step 400 "is similar to the previously described step 400. It comprises updating the correction factors jk2 associated with the parameter of position k2, as a function of the calibration function Fk2 associated with the value k2 of this second position parameter, according to the equality jk2 = argmax; [FK2 (j)]. The correction function J2, such that J2 (k2) = jk2, is then determined. The method described according to this embodiment can be generalized to the determination of N position parameters, the calibration method then comprising: the determination of N sets of correction factors, each set being associated with a position parameter , with and N 1, each set comprising as many correction factors jk, as values ki can be taken by each I 'position factor ki. In other words, according to this embodiment, N correction functions Jk ,, are determined, each correction function assigning a correction factor jk, as a function of the value ki taken by the said position parameter, according to the expression jk, = Jk, (ki). This determination is made for each correction parameter by repeating steps 100 to 600 previously described. the adjustment of the values of a correction set jk, by the following steps: irradiation of the detector with said reference source, with each interaction, determination of the values of the various of position parameters, and updating of a function Fki calibration associated with the parameter ki, according to the formula: For all j between and imax, F / (J) = Ffd 1 (i) Spre fj Gel j kn XA) where: - Fk, P-1 (j) is the value of the term Fk, (j) preceding the update, - Fk, P (j) is the value of the term Fk, (j) during the update, - A is the amplitude of the signal S measured by an electrode following each position parameter interaction k ,. In this example, we consider the signal SAO measured by the winning anode AO, so A = ASA °, -k correspond to the correction factors corresponding to the different values taken by the position parameters of rank n, with these factors being initialized. at an arbitrary value jkr, °, either previously determined. According to one variant, steps c), d) and e) are repeated iteratively, for the same value k i of said position parameter, each step e) resulting in the determination of a partial correction factor j i denoting the rank of the iteration associated with the value ki of the I 'position parameter. Following the first iteration, step d) includes updating each term Fki (j) by weighting said amplitude A by the product of the partial correction factors M ... jkqi, associated with the value ki of the I '. position parameter of the interaction, these partial correction factors having determined during the previous q iterations. It can be seen that, as iterations proceed, the partial correction factors gradually tend towards the value 1. The method then comprises, during step e) of the last iteration, the determination of the correction factor jk, associated to each value ki of the I 'position parameter according to the product of -Qi partial correction factors jki ... jki determined during said successive iterations, Q designating the number of iterations necessary relative to the value Here of the I' parameter of position. Typically, according to this variant, Qi is between 2 and 10. Thus, before the first iteration, the correction factor associated with the value Here, denoted j2i, is initialized to a predetermined value, for example equal to 1. Each iteration qi results in the determination of a partial correction factor j, each partial correction factor composing the intermediate correction factor 1-11 / 1qii I used as a weighting term of the amplitude measured during the next iteration , of rank cwi. During step e) of the last iteration, it is possible to determine that: ## EQU1 ## In other words, during step e) of the last iteration, the factor correction factor jk, associated with the value Here of the ith position parameter is determined as a function of the product of the partial correction factors corresponding to the same value Here of the ith position parameter, these partial correction factors having been determined during the previous iterations . For example, if a series designates the determination of partial correction factors by applying steps c) to e) previously described, in each series, step d) comprises an iterative update of the calibration function. Fk such that: For all j between kiwi and jrnak, where: Fft (J) = F, "° (j) Spre f jar it jjci x K x A) - Fk, 1) 4 (j) is the value of term Fk, (j) preceding the iteration bead, in the row series q (each series here represents an iteration allowing the determination of a partial correction factor, qi ki - Fkr (j) is the value of the term Fki ( j) after the iteration period, in the series of rank q, - A is the amplitude of the signal S measured by an electrode following an interaction whose position parameter is equal to k, in this example, we consider the SAO signal measured by the winning anode AO, so A = AA0, - p is the iteration index corresponding to the rank of the update of the function of calibration.
[0040] At the first update, p = 1, - - h / a is the partial correction factor associated with the value k of the position parameter, determined during the series /, with IS-K is a weighting factor, obtained from the product of the correction factors j1, or .qn of the product of the partial correction factors I kn1.kn associated with the values of the other position parameters of the interaction. More precisely, the quantity K comprises the product of weighting terms tr, each weighting term being associated with the value kr of each ni 'position parameter, with n # i, each term tr being equal to: the initial value of the correction factor j j, for example set to 1, or an intermediate value of the correction factor j corresponding to the product of each partial correction factor ./"1 kn x --- Jicn, or il / == gin / ktn, where qn denotes the rank of the iteration relative to the value kn, - the value jkr, of the correction factor associated with the value kr "if, for this value kn, 1 = Qn; 1 a number sufficient iterations qn = Qn has been reached, in which case jk, -, fIcn = III = Jkfl- The formula for updating the calibration function, according to this example, can then be: qi-1 qn ./ FPkfio ) = Ffi the qiu) + Sprefi (Flikii 11.11cn XA) 1 = 1 and 1 = 1 qn denoting the number of iterations reached for the determination of the correction Steps c) to e) are then renewed until a stop criterion is reached: it can be a predetermined number of series, a small difference between two partial correction factors jkqi and jq + 1 kt successive. Typically, according to this variant, the number Qi of iterations is between 2 and 10. The correction factors of each value k of each position parameter are then stored in the memory 50.
[0041] Whatever the embodiment, each correction factor jk, jk, associated with the value k, k, of a position parameter can be determined for a plurality of discrete values k, k, of said position parameter. By interpolation, it is possible to determine the correction factors associated with values lying between said discrete values. It can in particular be a linear interpolation. An example of operation of a detection device as shown schematically in FIG. 1 will now be described. When an interaction occurs in the detector material 11, it generates a signal on the cathode 12 and on a plurality of elementary anodes. adjacent 13x. The locating circuit 30 determines the winning anode AO as well as the secondary anode A1 as a function of the electronic signals transmitted by each processing circuit 23x associated with an elementary anode 13x. The locating circuit determines the ratio A1 / A0, according to which it determines a first position parameter 1 (1 = f (A1 / A0), representing the coordinates (XINT, VINT) of the interaction in the plane of the anodes elementary.
[0042] The locating circuit 30 then determines the ratio Ak / A0, corresponding to the amplitude of the cathode signal to the amplitude of the signal of the winning anode, according to which it determines a second position parameter k2 = g ( K / A0), representing the ZINT coordinate of the interaction between the elementary anodes 13x and the cathode 12. The correction circuit 60 searches, in the memory 50, the correction parameters jki and jk2 respectively associated with said values k1 and k2 , then corrects the value of the amplitude of the winning anode Ao by weighting it by the product jkix jk2. The corrected amplitude A * is then such that A * = jklx ik2 X AO. During the operating phase of the detector, the calibration circuit 40 is not implemented.

权利要求:
Claims (16)
[0001]
REVENDICATIONS1. Method for calibrating a device (1) for detecting ionizing radiation, the device (1) comprising a detector (10) comprising a semiconductor detector material (11) extending between two electrodes (E, E ') , E ", 12, 13, 13x), at least one electrode being capable of generating a signal S of amplitude A during an interaction of an ionizing radiation in said detector material: a) irradiation of the detector (10) at using a known reference source (Sref), so as to generate interactions in the detector material (11), b) selecting a plurality of correction coefficients (j) and determining reference spectra (Spref) ,), each reference spectrum being associated with a correction coefficient (j), each reference spectrum representing a distribution of the amplitude (A) of the signal (S) generated when the detector is irradiated by said reference source (Sref ), the amplitude (A) of the detected signal being po ndérée by said correction coefficient (j), c) detecting a plurality of interactions in the detector during said irradiation and, for each interaction, measuring at least one amplitude A of an electronic signal S generated by a electrode E and determination of the value k of at least one position parameter of the interaction as a function of said amplitude A, d) for a plurality of values k of said position parameter, determination of a calibration function Fk such that each calibration function Fk is associated with one of the said values k of the said position parameter, each calibration function Fk is defined on a plurality of correction coefficients j, each term Fk (j) of the said calibration function represents a probability associated with said correction coefficient j when the position parameter takes the value k, - each term Fk (j) being updated, for each interaction whose position parameter is equal to al at the value k at which said calibration function Fk is associated, by combining the amplitude A measured during this interaction with each reference spectrum Spred, e) Determining a correction factor jk associated with each value k of the position parameter by identifying a remarkable point of said calibration function Fk.
[0002]
2. Calibration method according to claim 1, wherein step d) comprises: initializing each calibration function Fk associated with each value k of said position parameter, ii each interaction detected, determining the value k of said position parameter and update of the corresponding calibration function Fk, by adding, at each term Fk (j) of this function, the value of the corrected reference spectrum Spref; (A) at said amplitude A, iii repeating step ii until a predefined stopping criterion is reached.
[0003]
3. Calibration method according to one of the preceding claims, wherein step e) comprises determining the correction factor jk associated with the value k of the position parameter by identifying the value of j maximizing the value of the function Fk in which case: jk = argmaxi [Fk (j)] [0004]
4. Calibration method according to one of the preceding claims, wherein: - steps c), d) and e) are repeated iteratively, for the same value k of said position parameter, each step e) leading to the determining a partial correction factor j, q designating the rank of the iteration, and in which, following the first iteration, step d) comprises updating each term Fk (j) by weighting said amplitude A by the product of the partial correction factors j -... jkq, associated with the value k of the position parameter of the interaction, these factors having determined during the previous iterations, the method then comprising, at the time of step e) of the last iteration, determining the correction factor jk associated with each value k of the position parameter as a function of the product of the partial correction factors determined during said successive iterations, Q denoting the number of iterations. ations.
[0005]
Calibration method according to claim 1 comprising: in step c), determining a set of N position parameters of the interaction as a function of the signal S, N being an integer greater than 1, step d), selecting a position parameter i from among the N position parameters, and, for a plurality of values Here of said ith position parameter, determining a calibration function Fki such that: - each function of Fki calibration is associated with one of said values Here said ieme position parameter, - each calibration function Fki is defined on a plurality of correction coefficients j, - each term Fki (j) of said calibration function represents a probability associated with said coefficient correction factor j when the position parameter takes the value ki, - each term Fki (j) being updated, for each interaction whose ieme position parameter is equal to the value ki at which said Calibration function Fki is associated, by weighting the amplitude A, measured during this interaction, by at least one correction factor j, said correction factor j, being determined as a function of the value kr, of at least one nor a position parameter characterizing the interaction, and by combining the amplitude A thus weighted with each Spred reference spectrum, in step e), the determination of the correction factor jki associated with each value Here of the i th position parameter according to said calibration function Fki associated with the ieme position parameter, by identifying a remarkable point of this function.
[0006]
6. Calibration method according to claim 5, wherein step d) comprises: the initialization of each calibration function Fki associated with each value Here of the ith position parameter Here, ii with each interaction detected: the weighting of the value of the amplitude A measured during this interaction by at least one correction factor jkr, depending on the value kr, of an nth position parameter characterizing the interaction, addition, to each term Fki (j) of said calibration function, the value of the Spref corrected reference spectrum; (A *), A * representing the value of the amplitude A thus weighted, iii repeating step ii until a stop criterion is reached.
[0007]
7. Calibration method according to one of claims 5 or 6, wherein step e) comprises determining the correction factor jk, associated with each value k, of the i 'position parameter by identifying the value of jk, maximizing the value of the function Fk ,, in which case: ha = argmaxi [Fia (j)] [0008]
8. Calibration method according to one of claims 5 to 7 wherein - steps c), d) and e) are repeated successively for each value ki of each of the N position parameters characterizing each interaction, - and in which l step d) comprises updating each term Fki (j) by weighting said amplitude A by the product of correction factors j1, each correction factor jkr being associated with the value kr, of a ni 'parameter of position, each correction factor jkr being either set at a predetermined initial value or determined during a previous iteration.
[0009]
9. Calibration method according to one of claims 5 to 8, wherein: - steps c), d) and e) are repeated iteratively, for the same value ki of said i 'position parameter, each step e ) leading to the determination of a partial correction factor jki, qi denoting the rank of the iteration, - and in which, following the first iteration, step d) comprises updating each term Fk, ( j) by weighting said amplitude A by the product of the partial correction factors M ... jkq associated with the value ki of the I 'position parameter of the interaction, these factors having determined in the course of the previous q, iterations then comprising, during step e) of the last iteration, the determination of the correction factor jki associated with each value Here of the position parameter as a function of the product of the partial correction factors determined during said successive iterations, C) :, designating the number of iterations.
[0010]
10. Calibration method according to claim 9, wherein during step d), the calibration function Fki is updated by also weighting the amplitude A by a weighting factor K, this factor being obtained from produces correction factors j1, or partial correction factors / kn. ". kn, associated with the values kr, of the other position parameters of the interaction.
[0011]
11. Calibration method according to one of the preceding claims, wherein step b) comprises: the loading of an initial reference spectrum Sore, previously stored in a memory, corresponding to said reference source Sref, this spectrum Sp f representing the distribution of the amplitude of the detected signal when the detector is irradiated by said reference source Sref, ii from the initial reference spectrum Sore, the determination of a plurality of reference spectra Sprefl, corresponding to a coefficient correction factor j, each reference spectrum Sore, being obtained from the initial reference spectrum Sore, each value Spref, (i) of the reference spectrum being obtained from the value Spref (i / j).
[0012]
12. Calibration method according to one of the preceding claims, wherein a position parameter k, k, of an interaction is obtained by combining signals resulting from said interaction, from different electrodes (12, 13, 13x).
[0013]
13. Calibration method according to one of the preceding claims, wherein the anode (13) of the detector being divided into a plurality of elementary anodes (13x) arranged in a plane, the value (k, k,) of a position parameter to an interaction is obtained by combining signals resulting from said interaction, said signals being generated, - by an elementary anode (AO) and by a cathode (12), or - by two elementary anodes (AO, al).
[0014]
The calibration method according to claim 13, wherein the anode of the detector is divided into a plurality of elementary anodes arranged in a plane, the value (k, k) of a position parameter of each interaction is obtained. by combining the signals: of the elementary anode having collected the maximum amplitude signal, called the winning anode (AO), and of the adjacent elementary anode of the winning anode having collected the maximum amplitude signal, referred to as the secondary anode (A1), in which case the position parameter determines the position of said interaction in the plane of said anodes.
[0015]
The calibration method according to claim 13, wherein the anode of the detector is divided into a plurality of elementary anodes, and wherein the value (k, k) of a position parameter of each interaction is obtained by combining the signals of the elementary anode having collected the maximum amplitude signal, called the winning anode (AO), from the cathode (12, K), in which case the position parameter determines the position of the interaction between the winning anode and the cathode, this position parameter then corresponding to a depth of detection.
[0016]
A device for detecting an ionizing radiation, comprising: a) a detector (10), comprising a semiconductor detector material (11), extending between two electrodes (12, 13, 13g), the detector being suitable to generate charge carriers under the effect of an interaction with said ionizing radiation, b) an electronic processing circuit (22, 23g), able to form a signal S from charge carriers collected by an electrode during an interaction and determining its amplitude A, c) an electronic positioning circuit (30), able to determine the value (k, k,) of at least one position parameter at the interaction in the material detector from the signal of at least one electrode, d) an electronic calibration circuit (40), configured to implement steps b), e) of one of the methods of claims 1 to 15.

类似技术:

---

公开号 | 公开日 | 专利标题

---

FR3030778A1|2016-06-24|METHOD FOR CALIBRATING A IONIZING RADIATION DETECTOR AND DEVICE THEREFOR

---

EP2486425B1|2020-04-01|Method for processing data from an ionizing-radiation detector

---

FR2953299A1|2011-06-03|METHOD FOR CALIBRATING AN X-RAY DETECTOR

---

EP2510338A1|2012-10-17|Method and device for recognizing a material using the transmission function thereof

---

EP0810631A1|1997-12-03|High resolution radiographic imaging device

---

CN104838287B|2018-08-17|Adaptive follow current for photon counting detector compensates

---

EP2904425B1|2019-05-01|Method and device for detecting ionising radiation using a pixellated photodetector

---

EP2434752B1|2018-01-10|Method for determining and correcting the response stability of a semi-conductor matrix detector

---

Xu et al.2019|Detection and analysis of laser driven proton beams by calibrated Gafchromic HD-V2 and MD-V3 radiochromic films

---

EP3237932B1|2019-08-21|Method for calibrating an ionising radiation detector and associated device

---

FR2702571A1|1994-09-16|Device for imaging ionizing particles by means of a multiwire proportional chamber.

---

FR3046241A1|2017-06-30|METHOD FOR CALIBRATING AN X-DIFFRACTION ANALYSIS SYSTEM

---

EP3674753B1|2021-10-06|Method for calibrating a spectrometry detector

---

FR3105444A1|2021-06-25|Method for determining a bias affecting pixels of a pixelated ionizing radiation detector

---

EP3629063A1|2020-04-01|Spectrometry system, associated spectrometry method and computer program product

---

EP3851837A1|2021-07-21|Method for establishing a spectral response function of a system for measuring by x or gamma ray

---

Viero et al.2015|HERMES: CURRENT COSMIC INFRARED BACKGROUND ESTIMATES ARE CONSISTENT WITH CORRELATED EMISSION FROM KNOWN GALAXIES AT Z< 4

---

FR2964195A1|2012-03-02|DIFFUSER FIRST SPECTRUM EXTRACTION METHOD, ASSOCIATED DEVICE AND COMPUTER PROGRAM

---

Stukachev et al.2011|Calibration method for spatial distribution of photosensitivity of matrix photodetectors

---

EP2368140A1|2011-09-28|Method for determining the spectral and spatial distribution of braking photons, and related device

---

FR3032063A1|2016-07-29|CONFIGURED ELECTRONIC CIRCUIT FOR CALIBRATING SILICON PHOTOMULTIPLIERS PHOTONLY DETECTING DEVICE, PHOTON DETECTING DEVICE, CALIBRATION METHOD, AND CORRESPONDING COMPUTER PROGRAM.

---

Suzuki et al.2011|Evaluation of characteristics of CdTe detector by laser pulses

---

Harrison et al.2010|A Study on the Effects of Aging on Radio Chromic Film

---

Morel1990|Measurement of the inclusive production cross section of J/Ψ in collisions proton-proton and antiproton-proton at√ S= 24.3 GeV

同族专利:

---

公开号 | 公开日

---

EP3237931A1|2017-11-01|

---

EP3237931B1|2018-11-21|

---

WO2016102404A1|2016-06-30|

---

EP3237931B8|2019-01-23|

---

FR3030778B1|2017-01-27|

---

US10132944B2|2018-11-20|

---

US20170350995A1|2017-12-07|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

FR2950979A1|2009-10-07|2011-04-08|Commissariat Energie Atomique|METHOD FOR PROCESSING DATA FROM A IONIZING RADIATION DETECTOR|

---

US20110297838A1|2010-06-04|2011-12-08|General Electric Company|Charge loss correction|

---

US6583420B1|2000-06-07|2003-06-24|Robert S. Nelson|Device and system for improved imaging in nuclear medicine and mammography|

---

FR3030778B1|2014-12-22|2017-01-27|Commissariat Energie Atomique|METHOD FOR CALIBRATING A IONIZING RADIATION DETECTOR AND DEVICE THEREFOR|

---

FR3030779B1|2014-12-22|2017-01-27|Commissariat Energie Atomique|METHOD FOR CALIBRATING A IONIZING RADIATION DETECTOR AND DEVICE THEREFOR|FR3030778B1|2014-12-22|2017-01-27|Commissariat Energie Atomique|METHOD FOR CALIBRATING A IONIZING RADIATION DETECTOR AND DEVICE THEREFOR|

---

US10481285B1|2018-08-13|2019-11-19|General Electric Company|Systems and methods for determination of depth of interaction|

---

US10976452B2|2018-08-13|2021-04-13|General Electric Medical Systems Israel, Ltd. |Systems and methods for improved medical imaging|

---

FR3091359B1|2018-12-26|2020-12-11|Commissariat Energie Atomique|Method for calibrating a spectrometric detector|

---

RU2736011C1|2019-10-01|2020-11-11|Федеральное Государственное Унитарное Предприятие "Всероссийский Научно-Исследовательский Институт Автоматики Им.Н.Л.Духова" |Radiation monitor and method of detecting pulsed neutron radiation|

---

US11092701B1|2020-07-07|2021-08-17|GE Precision Healthcare LLC|Systems and methods for improved medical imaging|

法律状态:
2015-12-31| PLFP| Fee payment|Year of fee payment: 2 |

---

2016-06-24| PLSC| Search report ready|Effective date: 20160624 |

---

2016-12-29| PLFP| Fee payment|Year of fee payment: 3 |

---

2018-01-02| PLFP| Fee payment|Year of fee payment: 4 |

---

2018-12-31| PLFP| Fee payment|Year of fee payment: 5 |

---

2020-10-16| ST| Notification of lapse|Effective date: 20200906 |

优先权:

---

申请号 | 申请日 | 专利标题

---

FR1463093A|FR3030778B1|2014-12-22|2014-12-22|METHOD FOR CALIBRATING A IONIZING RADIATION DETECTOR AND DEVICE THEREFOR|FR1463093A| FR3030778B1|2014-12-22|2014-12-22|METHOD FOR CALIBRATING A IONIZING RADIATION DETECTOR AND DEVICE THEREFOR|

---

EP15816169.5A| EP3237931B8|2014-12-22|2015-12-18|Method for calibrating an ionising radiation detector and associated device|

---

US15/538,494| US10132944B2|2014-12-22|2015-12-18|Method for calibrating an ionising radiation detector and associated device|

---

PCT/EP2015/080658| WO2016102404A1|2014-12-22|2015-12-18|Method for calibrating an ionising radiation detector and associated device|