Method, apparatus and medium for processing calibration spectra

Abstract

The present invention is a method for processing a calibration spectrum acquired by a spectroscopic detector of X-ray or gamma-ray photons. The method includes considering a parametric form of a calibration function, which relates the position of an energy channel to an energy value. The method includes comparing the channel of a peak of the calibration spectrum with the emission energy of a calibration isotope. This comparison allows the value of a parameter of the calibration function to be defined.

Description

[Technical field]

[0001] The technical field of the invention is spectroscopy applied to the detection of radiation.

[0002] prior art Devices for detecting ionizing radiation, based on gaseous detection materials, semiconductors or scintillators, are able to acquire electrical pulses formed by the interaction of the radiation in the detection material. The amplitude of each pulse depends on the energy deposited by the radiation during each interaction. These devices are usually connected to a spectrometric circuit. Such devices are able to acquire a spectrum, which corresponds to a histogram of the amplitudes detected during the irradiation period. The spectrum is formed according to a number of different amplitude channels, usually hundreds or thousands of channels. Each channel has a corresponding narrow amplitude band. Each channel is assigned a value corresponding to the number of pulses detected during the irradiation period, the amplitudes of which are within the amplitude band corresponding to the channel.

[0003] Spectroscopic measurement systems are now widely industrialized. The fields of application are wide and include, in particular, the measurement of nuclear waste, equipment or facilities, or environmental radiation monitoring. Software allows parameterization of the processing of the pulses, the driving and automatic interpretation of the measurements. Summary of the Invention [Problem to be solved by the invention]

[0004] However, some steps are difficult and leave ample room for manual setting. That is also the case, for example, for the energy calibration, which allows the establishment of a relationship between the amplitude of the pulse and the energy released by the interaction that generated the pulse. The relationship between amplitude and energy depends on the detector used and also on the parameterization of the spectroscopic measurement circuit, for example the amplification gain, the pulse shaping parameters or the number of channels. Moreover, for the same device, the relationship between amplitude and energy is prone to change unless the settings are changed, necessitating regular calibration.

[0005] Energy calibration is usually performed by exposing the detector to a calibration object whose emission energy is known. The calibration object may contain one or more isotopes whose emission energy is known. This calibration involves obtaining a spectrum of the photons emitted by the isotopes and then identifying the main peaks present in the spectrum. Some of the detected peaks correspond to known emission energies. Each detected peak is spread on either side of a central channel. Energy calibration allows the identification of different channel (i.e. the central channel of the peak)-emission energy pairs. From these different pairs, an analytical function, called a calibration function, is determined that relates the rank of each channel to the energy.

[0006] Energy calibration is a frequent operation that must be performed accurately to avoid misinterpretation of measurements. The invention described below facilitates automation and reproducibility of energy calibration. It also allows the determination of a calibration function with a maximized number of peaks considered. [Means for solving the problem]

[0007] A first subject of the invention is a method for processing a calibration spectrum produced by a spectroscopic measurement device, the spectroscopic measurement device comprising: a detector configured to detect particles and to form, at each detection, a pulse whose amplitude depends on the energy of the particle that has interacted in the detector; a spectrometry circuit configured to form a spectrum, the spectrum corresponding to the number of particles detected in a channel, each channel being assigned a rank, each rank having a corresponding pulse amplitude; The method comprises: a) positioning the spectroscopic measurement device so that it faces a calibration object which emits a number of particles with different known emission energies to a detector; - b) detecting a portion of the particles by a detector and forming a calibration spectrum of the detected particles, the calibration spectrum comprising a plurality of different peaks, each peak corresponding to a known energy; - c) assigning a channel rank to each detected peak; - d) determining a calibration function based on the ranks of the channels assigned to the different peaks in step c) and on the emission energies, the calibration function determining an energy based on the rank of each channel, the calibration function being a bijective function; Step d) is di) a sub-step of considering an analytical model of the calibration function, the analytical model being parameterized by a set of parameters; - d-ii) for several different values ​​of each parameter, applying a calibration function to the rank of each channel assigned to each detected peak, and obtaining, for each rank, an energy determined by the calibration function; or applying the inverse of the calibration function to each emission energy and obtaining, for each emission energy, a channel determined by the inverse function; - d-iii) a sub-step of determining a value for each parameter, the values ​​of the parameters for which the energy determined in d-ii) is closest to the emission energy of the calibration object; determining the value of a parameter such that the channel determined in step d-ii) is closest to the channel assigned to each detected peak; The present invention is characterized by comprising:

[0008] According to one embodiment, sub-step d-iii) comprises: - calculating the likelihood function for the value of each parameter; - It may involve estimating the value of each parameter that maximizes the likelihood function.

[0009] Step c) may include determining a width of each peak, and the likelihood function may take into account the width of each peak.

[0010] Sub-step d-iii) - following step c), considering a number of different pairs, each pair comprising an emission energy and a channel assigned to a peak, determining a value for each parameter that allows the calibration function to relate several pairs, each determined value forming an initial value for each parameter, It may include defining, for each initial value, a search range extending in the vicinity of said initial value, such that the value of each parameter that maximizes the likelihood function is estimated within a search range defined in the vicinity of the initial value of said parameter.

[0011] Sub-step d-iii) is preferentially carried out by an optimization algorithm. According to one possibility, step c) comprises selecting peaks of the calibration spectrum depending on a selection criterion for the peaks. The selection criterion may be the number of photons detected in each peak or a signal-to-noise ratio determined for each peak.

[0012] Sub-step d-iii) may comprise determining the value of at least one parameter based on a priori considerations regarding the value of said parameter.

[0013] The particles may be selected from photons, neutrons, and charged particles.

[0014] A second subject of the invention is a device adapted to obtain a spectrum of particles emitted by an object, said device comprising: a detector configured to detect particles and to form, at each detection, a pulse whose amplitude depends on the energy emitted by the particle that has interacted in the detector; a spectroscopic measurement circuit configured to form a spectrum corresponding to the distribution of amplitudes of the pulses detected by the detector; a processing unit programmed to carry out step d) of the method according to the first subject of the invention on the basis of the spectrum.

[0015] A third subject of the invention is a medium adapted to be connected to a computer, said medium comprising instructions for carrying out step d) of the method according to the first subject of the invention on the basis of a spectrum representative of the energy of the detected particles, said medium being integrated in the computer or being connected to the computer by a wired or wireless link.

[0016] A fourth subject of the invention is a method for processing a calibration spectrum produced by a spectroscopic measuring device, the spectroscopic measuring device comprising: a detector configured to detect particles and to form, at each detection, a pulse whose amplitude depends on the mass of the particle that has interacted in the detector; a spectrometry circuit configured to form a spectrum, the spectrum corresponding to a number of particles detected in a plurality of channels, each channel being assigned a rank, each rank corresponding to an amplitude of a pulse; The method comprises: a) exposing the spectroscopic measuring device to a calibration object containing particles of known mass; - b) detecting all or a portion of the particles by a detector and forming a calibration spectrum of the detected particles, the calibration spectrum comprising a number of different peaks, each peak corresponding to a known mass; - c) assigning to each peak a channel rank; - d) determining a calibration function based on the ranks of the channels assigned to the plurality of distinct peaks in step c) and the known distinct masses, the calibration function determining a mass from the rank of each channel, the calibration function being a bijective function; The method further comprises, in step d), di) considering an analytical model of the calibration function, parameterized by a set of parameters; - d-ii) for different values ​​of each parameter, applying a calibration function to the rank of each channel assigned to each peak, and obtaining, for each rank, a mass determined by the calibration function; or applying an inverse of the calibration function to each mass to obtain, for each mass, a channel determined by the inverse function; - d-iii) determining values ​​of parameters, The value of the parameter for which the energy determined in d-ii) is closest to the mass of the particle of the calibration object; determining the values ​​of the parameters for making the channel determined in d-ii) closest to the channel assigned to each peak; The present invention is characterized by comprising:

[0017] The invention will be better understood upon reading the description of the exemplary embodiments presented in the following description in conjunction with the figures listed below. [Brief description of the drawings]

[0018] [Figure 1] FIG. 1 shows an example of an apparatus that allows the implementation of the invention. [Figure 2A] FIG. 2A shows an example of a spectrum. [Figure 2B] FIG. 2B is a detail of FIG. 2A. [Diagram 3] FIG. 3 shows a spectral peak and the application of a second-order linear differential operator to the spectrum. [Figure 4] FIG. 4 shows the main steps of an embodiment of the method according to the invention. [Diagram 5] FIG. 5 shows the correlation between the emission energy and the central channel of the detected peak. [Figure 6] FIG. 6 shows the correlation between the center channel of the detected peak (X-axis) and the emission energy (Y-axis). [Figure 7A] Figure 7A shows a calibration spectrum acquired using a germanium detector, with the locations of the channels assigned to the different detection peaks indicated. [Figure 7B] Figure 7B shows a calibration spectrum acquired using a germanium detector, with the locations of the channels assigned to the different detection peaks indicated. [Figure 7C] Figure 7C shows the values ​​of the likelihood function of the parameters of the calibration function for different values ​​of said parameters, and corresponds to the spectra shown in Figures 7A and 7B. [Figure 7D] FIG. 7D is a detail of FIG. 7C. [Figure 7E] Figure 7E shows the value of the likelihood function of another parameter of the calibration function for different values ​​of said parameter, and corresponds to the spectrum shown in Figures 7A and 7B. [Figure 7F] FIG. 7F is a detail of FIG. 7E. [Figure 7G] FIG. 7G shows the correlation between the channel (X-axis) assigned to the detected peak and the emission energy (Y-axis) of the spectrum shown in FIG. 7A. [Figure 8A] Figure 8A shows a calibration spectrum acquired using a LaBr3 (lanthanum bromide) detector, along with the channel positions assigned to the different peaks detected and the range of uncertainty for those positions. [Figure 8B] Figure 8B shows the values ​​of the likelihood function of the parameters of the calibration function for a number of different values ​​of said parameters, and corresponds to the spectrum shown in Figure 8A. [Figure 8C] FIG. 8C is a detail of FIG. 8B. [Figure 8D] Figure 8D shows the value of the likelihood function of another parameter of the calibration function for a number of different values ​​of said parameter, and corresponds to the spectrum shown in Figure 8A. [Figure 8E] FIG. 8E is a detail of FIG. 8D. [Figure 8F]FIG. 8F shows the correlation between the channel (X-axis) assigned to the detected peak and the emission energy (Y-axis) of the spectrum shown in FIG. 8A. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] An apparatus 1 allowing the implementation of the invention is shown in Fig. 1. This apparatus is a measurement system and includes a detector 10 capable of interacting with radiation 5 emitted by an object 2. The object 2 is here nuclear waste, which contains a number of different radioisotopes 2 j Usually, the possible radioisotopes present in the object to be measured are known in advance. In particular, it is possible to create a list containing the radioisotopes potentially present in the object to be analyzed. The subscript j denotes each radioisotope.

[0020] In the example shown, the detector 10 is adapted to detect ionizing photon radiation. Ionizing photon radiation is understood to be, for example, X-type or gamma-type photon radiation formed by photons whose energy is between 1 keV and 2 MeV. The method according to the invention can also be applied to the detection of other particles, for example neutrons, charged particles (for example alpha or beta rays), or photons.

[0021] In the example shown, the detector comprises a semiconductor material of the germanium (Ge) type, but the semiconductor material may also be, for example, a semiconductor material of the Si, CdTe, CdZnTe type, commonly implemented for the detection of ionizing photons. The photons forming the incident radiation interact with the detector material, which is subjected to a bias voltage V. Each interaction generates charge carriers, which are collected by an electrode, usually the anode.

[0022] Other types of detectors may be used, for example a scintillator coupled to a photon / charge carrier converter or a gas detector of the ionization chamber type, provided that they allow the collection of an electric charge Q under the influence of the energy E released by the ionizing radiation during interaction in the detector 10. Exemplary common detectors of the scintillator type include NaI(Tl) or LaBr3.

[0023] The detector 10 is connected to an electronic circuit 12 configured to generate a pulse, the amplitude of which depends on and is preferably proportional to the amount of charge collected during the interaction, which corresponds to the energy deposited by the radiation during the interaction.

[0024] In the illustrated example, the detector 10 is connected to a cryostat 16, which contains liquid nitrogen that maintains the Ge detector at an operating temperature.

[0025] The electronic circuit 12 is connected to a spectroscopic unit 13 arranged downstream of said electronic circuit, which allows to collect together all the pulses formed during the acquisition period. Each pulse corresponds to an interaction of the incident radiation with the detection material. The spectroscopic circuit then classifies the pulses according to their amplitude A and provides a histogram containing the number of pulses detected according to their amplitude. This histogram is an amplitude spectrum. Usually it is obtained with a multi-channel analyzer. Each amplitude is discretized according to a channel, each channel being assigned an amplitude band. The value of each channel of the spectrum corresponds to the number of pulses whose amplitude is within the amplitude band assigned to the channel. Each amplitude band corresponds to an energy band, the correlation being bidirectional. Each channel is therefore assigned an energy band or an amplitude band.

[0026] The relationship between amplitude and energy can be established by illuminating the detector with a calibration object that emits radiation whose energy is known. This is in particular radiation with a known energy value and with at least one discontinuity or energy peak. This operation is usually referred to by the term energy calibration. In the case of gamma spectroscopy, the detector is 152 The detector is exposed to a calibrated source of Eu type, which generates photons with a known emission energy. It is also possible to implement other isotope sources, as described below in one of the exemplary embodiments. The isotope or isotopes preferably emit photons according to at least two different emission energies. The emission energies are spread between a minimum and a maximum. The difference between the minimum and maximum preferably covers the spectral region in which the detector is envisaged to be used.

[0027] The energy calibration is performed using the calibration function f e This allows the establishment of an analytical relationship between the measured amplitude and the energy. The calibration function f e Taking into account the above, by changing the variables, it is possible to assign channel values ​​to energy values, rather than amplitudes. Indeed, the amplitude depends on the detector and the settings, whereas the energy corresponds to a set physical quantity that is detector-independent.

[0028] The spectrum y is a histogram of the amplitudes of each pulse detected and discretized according to the energy or amplitude channel. Each channel is assigned an amplitude band, which becomes an energy band by utilizing a calibration function. The spectrum can be expressed in vector form (y1,...y k ,...y n ), where n corresponds to the total number of channels. Each channel is assigned a rank k, where 1≦k≦n. Each channel of rank k is assigned a low amplitude A k and high amplitude A k+1 and its amplitude is A k and A k+1If the detected pulse is between k and 1, then the detected pulse is assigned to a channel of rank k.

[0029] Low amplitude A of each channel k is the low energy e k Corresponding to the high amplitude A of each channel k+1 is high energy e k+1 The amplitude-energy correlation corresponds to the calibration function f e Therefore,

number

number

[0030] amount TIFF2024546345000004.tif1217 corresponds to the position of the center of the channel of rank k relative to the number of channels n of the spectrum. This is the standardized rank of each channel, lying between 1 (k=1) and (k=n). The standardization allows the establishment of a calibration function that is independent of the number of channels n, the latter of which can be parameterized.

[0031] The apparatus comprises a processing unit 14, which is programmed to execute the steps of the algorithm described in relation to figure 4. The processing unit 14 is connected to the spectroscopic unit 13 and receives therefrom each measured spectrum.

[0032] The purpose of gamma spectroscopy is to measure the isotopes present in object 2. j The objective of the present invention is to identify the isotopes and to estimate their respective activities. 152Figure 2B shows the spectrum y obtained from a germanium detector exposed to a Eu-containing source. The x-axis corresponds to the channel and the y-axis corresponds to the number of pulses whose amplitude corresponds to each channel. Figure 2B is a detailed view of Figure 2A and corresponds to the rectangular area depicted in Figure 2A.

[0033] The represented spectrum y comprises several peaks, each of which corresponds to the emission energy of an isotope present in the object 2. These peaks form useful information of each spectrum, from which it is possible to identify the isotopes and to quantify their respective activities. The spectrum also comprises a continuum, corresponding to the scattering of the photons in the detector or before they reach the detector. This continuum corresponds to the part of the spectrum at or between each peak. The spectrum also comprises a background noise component, which is represented by statistical fluctuations.

[0034] One of the routine steps of spectrum processing is the detection of each peak. In Fig. 2B, different peaks are shown. The X-axis corresponds to the amplitude channel and the Y-axis corresponds to the number of pulses detected in each channel. The processing unit can then determine the characteristics of the channel of each peak, for example the central channel in which the peak is centered. The central channel is usually the channel that contains the largest number of detected pulses.

[0035] Each peak is bounded by two channels, each of which is the lower limit of the peak. i and upper limit sup i As a result, each peak is assigned an order i in ascending order of amplitude. Each peak of order i has rank k of its central channel. i The latter is identified by the lower limit of the peak inf i and upper limit sup i is placed between.

[0036] One possibility is to manually define the boundaries of each peak. Detecting each peak and defining the boundaries of each peak can be done, for example, by2 This can also be done automatically by implementing the second derivative operator denoted by . Figure 3 shows a portion of a spectrum y corresponding to a peak. In Figure 3, d 2 The second-order differential operator denoted by (y) is also shown. d 2 The vertical lines corresponding to local minima in (y) allow identification of the channel of rank k corresponding to the center of the peak. This differentiation operator can be a third order Savitzky or Golay filter with window width 11.

[0037] FIG. 4 shows generally the main steps of the method for determining the calibration function, which steps are described below.

[0038] Step 100: Obtaining a calibration spectrum y. During this step, the detector is exposed to a calibration object, which typically contains one or more calibration isotopes whose emission energies are known.

[0039] Step 110: Finding peaks in the calibration spectrum and assigning a channel to each found peak. k1...k i ....k I is used to denote the rank of the channel assigned to each peak detected in the calibration spectrum. I denotes the number of peaks detected in the calibration spectrum y, which are the emission energies E′1...E′ of the calibration isotopes, respectively. j ....E' J Corresponds to.

[0040] Each peak in the calibration spectrum y has a full width at half maximum (FWHM) w i An example of the FWHM is shown in Figure 2B.

[0041] The FWHM is also determined in peak detection by conventional means. For example, when peak detection is performed using a Savitzky and Golay filter as described in connection with FIG. 3, the FWHM of each peak is determined by the function d2 The full width at half maximum (w) can be estimated by determining the number of channels between two zero crossings of (y). Figure 3 shows the full width at half maximum (w). This full width at half maximum (FWHM) can be determined by manual determination of the peak boundaries followed by other conventional spectroscopy software means.

[0042] Step 120: Calibration function f e Consideration of an analytical model of the calibration function. In this step, an analytical model of the calibration function is defined, for example a linear function or a polynomial of a given degree. This model is parameterized by a set of parameters β.

[0043] Each energy E corresponding to the detected peak i But rank k i It is assumed that the channel is located at the center of e i and e i+1 The pulses of energy between are collected together. Thus, Equation 3 is assumed.

number

number

[0044] The calibration function can be a linear function, where the set of parameters is β=(β0;β1),

number

[0045] In the case of a second order polynomial, the calibration function is of the type of equation 5'.

number

[0046] Calibration function f e is necessarily a full monoclinic function. Usually, the channels are sorted according to increasing amplitude, in which case the calibration function is an increasing function.

[0047] E i is the calibration function f e By using rank k i represents the energy corresponding to the center amplitude of the channel of e is rank k i The energy corresponding to each channel is generated.

[0048] The calibration object emits energy E′1...E′ j ...E′ J J corresponds to the number of emission energies. For example, if the amount of photons in a peak is too small, all peaks corresponding to the emission energies may not be detected by peak detection. The rank k1....k corresponding to a particular peak detected i ....k I is the specific emission energy E′1...E′ j ...E′ J Corresponds to.

[0049] Figure 5 shows the channels k1....k corresponding to each detected peak. i ....k I An example of (I=5) in ascending order (k i >k i-1 ) and the emission energies E′1...E′ of the isotopes j ...E′ J ascending order (E′ j >E' j+1 ) (J=7). The calibration function is Maximize the number of occurrences according to TIFF2024546345000009.tif1789.

[0050] In Fig. 5, each occurrence is indicated by an arrow. If two emitted energies are close together, they can be associated with channels of the same rank. In Fig. 5, this particular case is indicated by energies E1 and E2.

[0051] Step 130: Establishing Probability Density

[0052] As follows, each emitted energy E1...E j ...E J is a function f e -1 By applying, rank x1...x j ...x J 100 corresponding channels.

number

[0053] Notation f e -1 (E′ j ,β) is f e -1 We show that depends on β.

[0054] As mentioned above, the detected peaks have a particular spectral width w i and the rank of the channel is k i This leads to uncertainty about the energy E′ emitted by each j For the rank k of the channel in which the peak is detected, i can be considered as a variable that follows a given probability density g, for example a Gaussian distribution, as follows:

number

[0055] Starting from (7), we can write

number

[0056] All combinations k i ,E′ j are considered to have the same probability. Therefore,

number

[0057] Step 140: Establishing the likelihood function and the maximization function.

[0058] By considering (7), (8), and (9), we can write

number

[0059] From this, we derive the likelihood function for β, denoted as L(β|κ,σ,E′), as follows:

number

[0060] Parameters that maximize the likelihood function The vector of TIFF2024546345000017.tif1215 can be estimated as follows:

number

number

[0061] Vector TIFF2024546345000020.tif1215 is the vector that maximizes the number of correlations between E′ and κ. The maximum occurrence value is obtained according to TIFF2024546345000021.tif1283.

[0062] The vector thus obtained TIFF2024546345000022.tif1215 uses the calibration function f e can be defined.

[0063] Figure 6 shows the calibration function f e This shows the situation where is a linear function. The calibration function f e is for each channel of rank k as follows:

number

[0064] On the other hand, energy E k corresponds to a channel of rank k, as follows:

number

[0065] Equation (13) becomes:

number

[0066] The optimization algorithm expressed in (16) is advantageous because it can be implemented near given values ​​of β0 and β1. This allows the use of a maximization algorithm around given initial values ​​of β0 and β1 within a given search range. This allows the avoidance of the "optimization trap" corresponding to the determination of a local maximum of β0 and β1. This also allows the optimization of the computation time. A search range is defined around each pair of values ​​β0 and β1.

[0067] In general, regardless of the type of calibration function, the vector TIFF2024546345000026.tif1215 is advantageous because it is determined by utilizing an optimization algorithm, and the vector We can solve (16) based on limited search ranges near the discrete initial values ​​of TIFF2024546345000027.tif1215, which correspond to continuous intervals extending near each initial value.

[0068] The definition of a limited search scope is explained below in conjunction with FIG.

[0069] In FIG. 6, the X-axis corresponds to the normalized rank k / n of each channel. The Y-axis corresponds to the highest emitted energy E′ j,max The emitted energy E′ is normalized by j corresponds to E′ j,max =E'4. Emitted energy is detected when a channel is assigned to it.

[0070] Channel k identified by searching for a peak in the calibration spectrum i is represented by the black vertical line. As mentioned before, the energy channel is iis assigned and is represented by two thin vertical lines on either side of a black vertical line. TIFF2024546345000028.tif1418 corresponds to a horizontal line. These are exact values ​​and have no associated uncertainties.

[0071] Taking into account the associated uncertainties, several lines are plotted in grey, which represent the standardized emitted energy TIFF2024546345000029.tif1418 and detection channel k i At least two intersections between Corresponds to TIFF2024546345000030.tif1327. It is taken into account that the calibration function is necessarily increasing when the ranks of the channels are arranged in ascending order of amplitude. Therefore, only lines that can correspond to increasing functions (β1>0) are plotted. The lines plotted in black are the lines that show the maximum number of intersections, in this case (E′1,k1), (E′3,k3) and (E′4,k4).

[0072] Figure 6 shows the definition of a restricted search space for implementing the optimization algorithm described above. It is based on the use of a bijective calibration function, E′ j (or TIFF2024546345000031.tif1418) and f e (k i ) and conversely, k i and f e -1 (E′ j ) to generate an emission energy and a rank of the channel assigned to the detected peak. The optimization algorithm described in (16) determines the initial values ​​of the parameters of the calibration function, which generates an emission energy and a rank of the channel assigned to the detected peak. The calibration function is performed near the initial value of TIFF2024546345000032.tif1215, so that the calibration function is performed for at least two pairs whose respective energies and ranks differ from one pair to the other, in other words, for two pairs (E′ j ;k i ) and (E′ j′ ;k′ i′ ), E′ j =f e (k j ), which confirms that The file is TIFF2024546345000033.tif1149.

[0073] From FIG. 6, it is possible to establish different initial values ​​for β0 and β1.

number

number

[0074] According to a variant, steps 130 and 140 are performed to determine the strongest emission energy E′ j and / or channel k i This is done by selecting the peaks that are considered to have the highest surface or have the highest signal-to-noise ratio. This allows the calculation time to be optimized. According to this variant, the number of emitted energies and the number of channels can be limited to a few tens. For example, I=30 and J=20. More generally, the method involves considering a selection criterion for each peak. The peaks that are processed in steps 130 and 140 are those that satisfy the selection criterion.

[0075] According to a variant, in step 140, an a priori is taken into account for the values ​​of one or more parameters constituting the set of parameters β. For example, a variation range is defined for one or more parameters. Each parameter considered a priori must be within a given variation range. The a priori consideration makes it possible to avoid the optimization algorithm generating unrealistic parameter values. The a priori is a constraint for the execution of the optimization algorithm, which is in addition to the condition that the calibration function is monotonic.

[0076] Experimental Testing The inventors carried out the above-mentioned calibration method using two types of detectors, namely, a Ge (germanium) type detector and a LaBr3 (lanthanum bromide) type detector, as shown in Fig. 1. Here, the results of the calibration are reported.

[0077] In the first series of tests, the energy calibration of the germanium detector was carried out. 152 A standard source of Eu was used. Calibration spectra were acquired for 3 hours. The source was placed at a distance of 40 cm from the detector. Spectra were acquired in 8192 channels. The calibration function was considered to be a linear function.

[0078] Figures 7A and 7B show the channels resulting from the peak detection algorithm. It can be seen that there are no peaks detected beyond the channel with rank 6600. In Figures 7A and 7B, the X-axis corresponds to the channel and the Y-axis corresponds to the number (i.e., number) of pulses detected and assigned to each channel. This is a conventional spectral representation.

[0079] The detection of the peaks made it possible to obtain 179700 pairs of possible channel correlations between the detected peaks and the emission energies. By taking into account that the calibration function is an increasing function, 82650 possible pairs are obtained. This corresponds to J=20 (number of emission energies considered) and I=30 (number of detected peaks), and the number of possible pairs N pairings is equal to

number

[0080] Figure 7C shows the value of the likelihood function L (Y-axis) as a function of β0 (X-axis). Figure 7D is a detail of Figure 7B. The value of β0 that maximizes this likelihood function is equal to 0.3689 keV.

[0081] Figure 7E shows the value of the likelihood function L (Y-axis) as a function of β1 (X-axis). Figure 7F is a detail of Figure 7E. The value of β1 that maximizes this likelihood function is equal to 1907.2475 keV.

[0082] FIG. 7G is a representation similar to FIG. 6 described above. The X-axis corresponds to the rank of the channel of the detected peak and the associated uncertainty. The Y-axis corresponds to the 152 corresponds to the emission energy of Eu (keV). By utilizing a calibration function that considers optimal values ​​of β0 and β1, the number of correlations between the channel and emission energy of the detected peaks can be maximized.

[0083] In the second series of tests, an energy calibration of the LaBr3 detector was performed. 22 Na, 57 Co, 60 Co and 133 A standard source of Ba was used. The main emission energies were 276.4 keV, 302.85 keV, 356.01 keV, 383.85 keV, 511 keV, 1173.23 keV, 1274.54 keV, and 1332.49 keV. Spectra were acquired in 4096 channels. The calibration function was assumed to be a linear function.

[0084] Figure 8A shows the channel obtained from the peak detection algorithm. Eight peaks were detected, which corresponds to the number of energy emissions. The detection of these peaks allowed us to obtain 1568 points by taking into account that the calibration function is an increasing function.

[0085] Figure 8B shows the value of the likelihood function L (Y-axis) as a function of β0 (X-axis). Figure 8C is a detail of Figure 8B. The value of β0 that maximizes this likelihood function is equal to -120.7273 keV.

[0086] Figure 8D shows the value of the likelihood function L (Y-axis) as a function of β1 (X-axis). Figure 8E is a detail of Figure 8D. The value of β1 that maximizes this likelihood function is equal to 1607.1911 keV.

[0087] FIG. 8F is similar to FIG. 7F and FIG. 6 described above. The X-axis corresponds to the rank of the channel of the detected peak and the associated uncertainty. The Y-axis corresponds to the emission energy as described above. By utilizing a calibration function that takes into account the values ​​of β0 and β1, the number of correlations between the channel of the detected peak and the emission energy can be maximized.

[0088] The calibration function may be different from a linear function, for example the calibration function may be a polynomial of degree 2 or higher.

[0089] The invention has been described in the context of gamma ray spectroscopy. The invention can be generalized to the spectroscopy of other types of photons or other types of particle radiation, in particular neutrons or charged particles. The invention can also be implemented in the detection of non-ionizing photons, for example photons in the infrared spectral range, photons in the visible spectral range, or photons in the near ultraviolet spectral range.

[0090] It will be appreciated that the invention is applicable to mass spectrometry, where the spectroscopic detector is configured to form a pulse whose amplitude depends on the mass of the particle that has interacted in the detector, and the establishment of a calibration function is performed by exposing the detector to a calibration object that includes a number of particles of known different masses, and the calibration function allows a bijective association between the amplitude of each pulse and the mass of the particle detected by the detector.

Claims

1. 1. A method for processing a calibration spectrum (y) formed by a spectroscopic measurement device (1), the method comprising: a detector (10) configured to detect particles and to form at each detection a pulse whose amplitude depends on the energy of the particle that has interacted in the detector; a spectroscopic measurement circuit (12, 13) configured to form a spectrum corresponding to the number of particles detected in a plurality of channels, each channel being assigned a rank and a pulse amplitude corresponding to each rank; The method comprises: -a) Different known emission energies (E') ｊ positioning the spectroscopic measurement device so that it faces a calibration object (2) that emits a plurality of particles having a .gtoreq..gtoreq..times ... - b) detecting a portion of the plurality of particles with the detector and forming a calibration spectrum (y) of the detected particles, the calibration spectrum comprising a plurality of distinct peaks, each peak corresponding to one of a plurality of known emission energies; -c) Detecting several peaks in the calibration spectrum and assigning a channel rank (k ｉ ), -d) from the ranks of the channels assigned to the different peaks in step c) and the emission energies (E' ｊ ), determining a calibration function, said calibration function determining an energy from said rank of each channel, said calibration function being a bijective function; Step d) is -d-i) the calibration function (f ｅ generating an analytical model of the equation (β), said analytical model being parameterized by a set of parameters (β); -d-ii) for each different value of said parameter of said set of parameters, - applying said calibration function to the rank of each channel assigned to each detected peak, to obtain, for each rank, an energy determined by said calibration function; or applying the inverse of said calibration function to each emission energy to obtain, for each emission energy, a channel determined by said inverse function; - d-iii) a sub-step of determining the value of each parameter, determining the values ​​of the parameters for which the energy determined in d-ii) is closest to the emitted energy of the calibration object; Or, a substep of determining the value of a parameter that makes the channel determined in d-ii) closest to the channel assigned to each detected peak. Including, The sub-step d-iii) comprises: - calculating the likelihood function for the value of each parameter L(β|κ, σ, E′) of said set of parameters; - each parameter that maximizes the likelihood function estimating the value of method.

2. - said step c) comprises determining the width (w i ) of each peak; the likelihood function is defined using the width of each peak, The method of claim 1.

3. The sub-step d-iii) comprises: - following step c), generating a number of different pairs ((E'j; k i )), each pair comprising an emission energy and a channel assigned to a peak; determining a value for each parameter of said set of parameters that allows said calibration function to relate several pairs, each determined value forming an initial value for each parameter; - defining, for each initial value, a search range extending near the initial value of each parameter, such that the value of each parameter that maximizes the likelihood function is estimated in the defined search range near the initial value of the parameter.

4. A method according to claim 1 or 2, wherein sub-step d-iii) is performed by an optimization algorithm.

5. A method as described in claim 1 or 2, wherein step c) includes selecting peaks in the calibration spectrum based on peak selection criteria.

6. The method described in claim 5, wherein the selection criterion is the number of photons detected at each peak or the signal-to-noise ratio determined for each peak.

7. A method according to claim 1, wherein sub-step d-iii) comprises determining the value of at least one parameter using a priori assumptions regarding the value of the parameter.

8. A method according to any one of claims 1 or 2, wherein the particles are selected from photons, neutrons, or charged particles.

9. An apparatus (1) configured to acquire a spectrum of particles emitted from an object, comprising: a detector (10) configured to detect particles and to form at each detection a pulse whose amplitude depends on the energy emitted by the particle that has interacted in the detector; a spectroscopic measurement circuit (12) configured to form a spectrum (y) corresponding to the distribution of amplitudes of the pulses detected by said detector; a processing unit (14) programmed to carry out, from said spectrum, said step d) of the method according to claim 1 or 2; An apparatus (1) comprising:

10. A medium configured to be connected to a computer, containing instructions for performing step d) of the method described in any one of claims 1 or 2 from a spectrum representing the energy of the detected particles.

Images