FR3090901A1 - METHOD FOR COUNTING PHOTONS ASSOCIATED WITH A RADIATION SOURCE - Google Patents

Abstract

The invention relates to a method for determining a share of photon count rate attributable to a photon source, from a radiation spectrum, the method comprising the following steps: detection (200) of a peak radiation, determination (300) of a first energy level and of a second energy level, these two levels delimiting a first energy window which includes the radiation peak, determination (400) of a high level , and preferably at a low level, respectively delimiting a high energy window and, where appropriate, a low energy window, calculation (500) of a part of the total counting rate of photons in the first window of energy which is attributable to a source of photons emitting according to the radiation peak, by subtracting from said total count rate the total count rate in the high energy window and possibly the total count rate in the low energy window. Figure 2

Description

Description

Title of the invention: METHOD FOR COUNTING PHOTONS ASSOCIATED WITH A SOURCE OF RADIATION

FIELD OF THE INVENTION AND STATE OF THE ART

The invention is in the field of acquisition and processing of spectrometric data for the characterization and localization of sources of photon emission, and in particular of radioactive or X-ray sources.

[0002] Techniques are known for real-time localization of a photon source, for example buried in a ground, by acquisition of photon emission spectrum data, with possibly application of methods of amplification of the weak signals. A photon detector is used for this purpose, which can be fixed or mobile. Such a detector is typically configured to detect gamma rays or X-rays, and is capable of providing a count rate of photons received per unit of time, for several given energy levels. The counting rate for an energy level corresponds to an amount of events detected per unit of time, the events detected being impacts of photons having said energy level. The counting rates and associated energy levels are used to characterize the energy sources present in the field. The measurement of the photon counting rate makes it possible to calculate the activity of a photon source present in the ground and positioned so as to emit photons in the direction of the detector head.

It is known to grid a surface on the ground at a plurality of measurement points, and to carry out measurements of photon counting rate at each of said points, with possibly a correlation between the measurements made at neighboring points, to locate sources of photons in the ground and perform a mapping.

However, known radiation detectors do not discriminate among unitary photon detection events, also known as "shots", between shots which are linked to the presence of a radioactive source, and linked shots to the natural activity of the land, forming a “background noise”. Typically, on an energy spectrum acquired at a measurement point on the ground, a radiation peak can be identified for an energy line to which a source emits, but the strokes detected for said energy line and lines d neighboring energy on the spectrum is not only due to the source.

The subsequent determination of the activity of the source is therefore skewed. GENERAL PRESENTATION OF THE INVENTION

There is thus a need for a method for determining photon counting rates which makes it possible to better isolate the activity of a source of photons emitting at a precise energy level, with respect to the background noise which n is not characteristic of the presence of a source, even when the contribution of the source is low compared to the background noise.

There is also a need for a method of determining the counting rate which is rapid and precise.

As such, the invention provides a method for determining a share of photon count rate attributable to a photon source, from a radiation spectrum comprising a photon count rate associated with each level of a plurality of energy levels,

The method comprising the following steps:

Detection of a local maximum count rate, said maximum constituting a radiation peak,

Determination of a first energy level and a second energy level, these two levels delimiting a first energy window which includes the radiation peak,

Determination of a high level higher than the second level, delimiting with the second level a high energy window, the high level being determined as a function of a region of interest of the first energy window,

And, preferably, determination of a low level lower than the first level, delimiting with the first level a low energy window, the low level then being determined as a function of the region of interest of the first window energy,

Calculating a share of the total count rate of photons in the first energy window which is attributable to a source of photons emitting according to the radiation peak, by subtracting from said total count rate the total count rate in the high energy window, and possibly the total count rate in the low energy window.

An advantage of the method of the invention is to allow the taking into account of the background noise. The number of strokes detected due to background noise for the energy level corresponding to the radiation peak identified in the spectrum is obtained from the number of total strokes detected for the high energy window, and if necessary for the window low energy, said windows being adjacent to the first energy window containing the peak.

One can thus obtain a photon count rate which is actually attributable to the source, independently of the background noise.

Recent technical developments having provided radiation detectors having an increased sensitivity to "hits", it is possible to scan more quickly a plurality of points on the ground to search for sources, without losing precision in detecting the counting rate.

The method of the invention also has the advantage of being simple to implement by a processing unit. The process can be integrated into a context of real-time localization of radioactive sources in an area and mapping of environmental radioactivity, for a crisis situation or not.

A method according to the invention may have the following characteristics, which are not limiting, taken alone or in any of the technically possible combinations:

- The method comprises determining the low level and the low energy window, the calculation step comprising subtracting the total count rate in the low energy window from the total count rate in the first window energy, the region of interest of the first energy window is defined using the width at half height of the detected radiation peak;

- the region of interest of the high energy window is equal to the region of interest of the low energy window;

- the regions of interest of the high and low energy windows are both equal to half of the region of interest of the first energy window;

- for the determination of the share of the total count rate attributable to the source, the total count rates are subtracted from the total count rates in the high energy window and in the low energy window;

- during the detection step, two radiation peaks associated with two distinct energy levels are detected in the spectrum, the first energy window comprising the two said peaks;

- the region of interest of the high energy window is equal to the region of interest of the first energy window;

- for the determination of the share of the count rate attributable to the source, we only subtract the total count rate in the high energy window;

- The method comprises a subsequent step of determining an activity of a photon source which would be located at the geographic coordinates of the measurement point;

- The method further comprises a preliminary step of acquiring the radiation spectrum at the measurement point.

The invention secondly targets a photon source identification system, comprising:

A radiation detector configured to measure a radiation spectrum, a processing unit and a memory associated with the processing unit, the detector being configured to transmit spectrum measurements to the processing unit,

The processing unit being configured to determine, from a spectrum acquired by the detector, a share of photon count rate attributable to a photon source by the implementation of a method for determining counting rate as defined above.

The system of the invention can, optionally and without limitation, have the following characteristics taken alone or in combination:

- the system is placed on a transport vector, the system being adapted to be moved over the surface of a ground, the radiation detector being adapted to acquire radiation spectra at a plurality of points on the surface;

- The system further comprises a support structure of the detector, and a propulsion unit adapted to suspend the support structure above the surface of a ground, the radiation detector being adapted to acquire spectra of radiation at a plurality of points on the surface;

- The radiation detector is not collimated;

- The radiation detector is configured to acquire gamma radiation spectra.

In another aspect, the invention relates to a computer program product comprising code instructions for implementing the above method when said instructions are executed by a processing unit.

GENERAL PRESENTATION OF THE FIGURES

Other characteristics, objects and advantages of the invention will emerge from the description which follows, which is given by way of nonlimiting example, accompanied by the following appended drawings among which:

[Fig.l]

Figure 1 schematically shows a photon source identification system of the invention.

[Fig.2]

Figure 2 illustrates the steps of a method for determining photon counting rates according to one embodiment.

[Fig.3a]

[Fig.3b]

[Fig.3c]

Figures 3a, 3b and 3c show theoretical examples of radiation spectra (shots per second depending on the energy level), corresponding respectively to background noise, to the energy emitted by a radionuclide buried on the ground , and the cumulative energy of said noise and said radionuclide.

[Fig.4a]

Figure 4a shows a first spectrum of real gamma radiation (counts per second depending on the energy level) acquired over a period of ten minutes.

[Fig.4b]

[Fig.4c]

[Fig.4d]

Figures 4b, 4c and 4d illustrate the implementation of certain successive steps of a counting rate determination method according to Ligure 2 from a second gamma radiation spectrum acquired over a period of one second.

[Fig.5a]

[Fig.5b]

[Fig.5c]

Figures 5a, 5b and 5c illustrate the implementation of certain successive steps for the determination of a count rate from a third spectrum of real gamma radiation acquired over a period of ten minutes.

DETAILED DESCRIPTION OF EMBODIMENTS

In what follows, the term "photon source" can designate a radioactive source, emitting by atomic de-excitation of gamma radiation and energy, or a source emitting infrared radiation, or even X-rays. follows in the case of a radioactive source emitting gamma photons.

The “count rate” or “photon count rate” designates a number of events of detection, by a radiation detector, of gamma photons in a certain energy line, per unit of time. We usually speak of "hits" to designate detection events. The counting rate can for example be given in counts per second.

We speak of "region of interest (or ROI)" to designate the width of an energy window, an energy window can be defined as a set of contiguous energy lines. Finally, we will speak of a "radiation peak" for a local maximum counting rate in a radiation spectrum.

System for identifying photon sources

Referring to Figure 1, a photon source detection system 1 comprises a transport vector 10. The transport vector is here an airborne vehicle such as a helicopter. Alternatively, the transport vector is a car, an airplane, a hand truck, etc., on which one or more radiation detectors are mounted. The transport vector can traverse the surface of a ground, allowing to traverse several measurement zones in succession on the ground surface. Here, the detector is configured to acquire and record gamma ray spectra.

In the case of the location of a radioactive source, the detector is advantageously a Nal detector, since it has a high sensitivity, which makes it possible to acquire measurements of one second which are representative of the area measured. In addition, no cooling is expected with this type of detector.

For the characterization of zones and the measurement of activities and / or location of a source (which will not be described in detail in the present description), the distance between the detector and a measurement point located at soil is a parameter to take into account in order to correct the measurement and bring it back to the ground surface.

The system 1 further comprises a memory 13 for storing the measurement results. Optionally, the system 1 also includes a calculation and processing unit 12. The detector is then provided with an interface 15 enabling it to transmit data to the calculation and processing unit.

The system may further comprise a positioning device, for example a satellite geo-positioning system (GPS), connected to the calculation and processing unit 12, to associate with the spectrometric data acquired at a point of measures the geographic coordinates of said point.

Finally, the system can, optionally, be provided with a temperature sensor capable of providing the processing unit 12 with a temperature measurement of the detector.

Alternatively, the system 1 does not include a calculation unit configured to perform the analysis of the measurement data. The memory 13 of the system 1 can then be connected via a communication network to a server configured to perform the data analysis.

Method for determining the count rate for a source

FIG. 2 shows the steps of a method 20 for determining, from a radiation spectrum acquired at a measurement point, a share of photon count rate attributable to a source.

The method 20 is for example implemented by a processing unit of the system 1 described above, said processing unit having been preconfigured with computer code instructions allowing the implementation of the method, or else implemented work by a remote server.

Optionally, a user of the system 1 can, via an interface, configure the analyzes carried out on each unit spectrum to target a given energy window. For example, if an operator is looking for a source of radioactive cesium, he can select an energy window including the 661.6 keV energy line, the cesium 137 decay energy being 661.6 keV. The operator can also select other analysis parameters such as other regions of interest, peak detection, artificial radionuclide indicators. It is also possible to extract various information related to the detector and to the GPS such as the elevation compared to the ground, the type of detector, the height.

In the context of the method 20 for determining counting rates, it is sought to eliminate the contribution of noises of two types: a first noise qualified as “background noise” and a second noise qualified as “radionuclide noise”.

In fact, the counting rate at the level of the radiation peak, in the energy window targeted by the user of the detector, generally does not correspond only to photons emitted directly, that is to say without dissemination, by the same source to be identified. Background noise and radionuclide noise are generally superimposed on these directly emitted photons. For the energy line corresponding to the peak, certain events correspond to the background noise linked to the natural activity of the terrain, still others may correspond to other sources emitting at higher energy levels, but having undergone Compton scattering and being detected at lower energy levels than their emission energy.

The principle of eliminating measurement noise is explained in relation to

Figures 3a to 3c.

The spectrum of Figure 3a corresponds to the background noise, the spectrum of Figure 3b corresponds to the cumulative activity related to the radiation peak at around 520 keV of a radionuclide and the noise of said radionuclide, and the spectrum of Figure 3c corresponds to the addition of the two spectra of Figures 3a and 3b.

It will be understood that the radiation detector is not necessarily configured to discriminate background noise, the noise of the radionuclide and the activity linked to the radiation peak. The detector therefore measures the spectrum of Figure 3c.

In these three figures, the left part gives a radiation spectrum materialized by successive points. The number of strokes per second is given on the ordinate and the energy is given in keV on the abscissa. Also shown on the left of the three said figures is a regression curve superimposed on the radiation spectrum, in dotted lines.

The right part of the three said figures represents only said regression curves on which a first window corresponding to a region of interest has been identified, as well as low and high energy windows.

In FIG. 3a, the background noise DI of the spectrum Cl corresponds to a linear regression curve.

Thus, if the region of interest is L, the sum of the area 30 under the curve DI on the low energy window of amplitude L / 2 and the area 32 under the curve DI on the window high energy amplitude L / 2 is equal to the area 31 under the curve DI in the region of interest.

In FIG. 3b, it is possible to extract from the spectrum C2 a first regression curve D2a corresponding to the peak and a second regression curve D2b corresponding to the noise of the radionuclide.

Note that the curve D2b has symmetry with respect to the vertical axis associated with the energy of the peak (about 520 keV). Thus, the area 40 under the curve D2b in the low energy window is equal to the sum of the areas 41 and 42 under the curve D2b in the region of interest.

Ligure 3c represents a spectrum C3 which is the sum of the spectra C1 and C2 and which corresponds to the spectrum observed by the detector. The regression curve D3b is the sum of the regression curves DI and D2b.

In view of the above, it is equivalent to subtract area 51 from the curve D3a in the region of interest, and to subtract areas 50 and 52 which correspond respectively to the low and high energy windows.

In contrast, areas 50 and 52 are directly observable in the measured radiation spectrum, unlike area 51.

During method 20, it is thus proposed to obtain, from a total count rate on a region of interest, the count rate attributable to a source giving rise to a radiation peak, in subtracting from said total count rate the count rates observed on high energy, and possibly low energy windows.

Returning to Ligure 2, in a step 100, a measurement of the radiation spectrum S is carried out at a measurement point by the detector 11. The acquired measurements are transmitted to the processing unit 12.

The acquisition of a radiation spectrum takes place for example during a calibration, in order to validate the response function of the detector to incident radiation, as a function of measurements provided by irradiations from sources with known activities. Indeed, each detector has its absolute efficiency (also called detector efficiency) and the calculation carried out to obtain the activity of radionuclides depends on said efficiency and varies according to the energy of said radionuclides.

Alternatively, the measurement is acquired in step 100 during a measurement campaign, in order to characterize a radiation source and / or assess a distribution of source (s) in a contaminated territory.

The geographic coordinates of the measurement point are also transmitted to the processing unit, these coordinates being able to be acquired by a map system.

Next, the processing unit performs steps 200 to 600 below based on the S spectrum at a point.

In a step 200, a radiation peak P is determined on the spectrum S. This peak corresponds to an energy line. The radiation peak can be detected automatically, by known methods, on the S spectrum (by detection of a local maximum of count rate). Alternatively, the energy line associated with the radiation peak can be predetermined, in particular if an atomic nature of the radioactive sources sought is already known.

FIG. 4a shows a measured radiation spectrum. This spectrum re9 presents a counting rate per photon in counts per second on the ordinate, and energy levels in kiloelectronvolt (or keV) on the abscissa. This spectrum was acquired over a period of ten minutes.

FIG. 4b shows a spectrum S of radiation acquired over a period of one second. This spectrum has a lower counting statistic than the spectrum in Figure 4a. This spectrum is for example acquired by the detector 1 represented in FIG. 1. The range of the abscissae is between 0 and 1000 keV and is more restricted than the range of the graph in FIG. 4a.

On the spectrum S, a step 200 of peak detection was carried out. A peak P is detected with the energy line at 513 keV. This peak constitutes a local maximum count rate.

In a step 300, a first and a second energy level are determined, delimiting a first energy window containing the energy line corresponding to the radiation peak. These energy levels are determined to cover a significant proportion of the detection events linked to the presence of the source to be identified. For example, the region of interest of the first energy window is defined using the width at half-height of the peak P of detected radiation.

In Figure 4c, there is shown a first energy window B _P around the peak P determined at the end of step 300, the first window being delimited by the energy levels Ni and N ₂ . The width for the peak P is 109 keV. The energy levels Ni and N ₂ are taken respectively at 458.5 keV and 567.5 keV.

The window B _P advantageously corresponds to a region of interest equal to the width at half height of the peak P.

The quantity of events detected in the energy window thus determined is useful for determining the activity of the source to be identified.

However, not all events are directly attributable to said source, according to the principles described above in relation to Figures 3a to 3c.

To subsequently evaluate the activity of the source to be identified, it is desirable to give an estimate of the share of the count rate, on the first energy window B _P , which is attributable to the source taken in isolation, in eliminating contributions from other sources and / or background noise.

In a step 400, the processing unit determines a third energy level N ₃ forming with the level Ni a low energy window B _b corresponding to energies lower than the energy line of the peak P, as well as a fourth level N ₄ forming with level N ₂ a high energy window B _H corresponding to energies higher than the energy line of the peak P. The window B _P is therefore framed by the windows B b θί B _h .

The levels N ₃ and N ₄ are determined as a function of the region of interest of the main window B _P.

Preferably, these levels are determined so that the region of interest of the windows B _b and B _H corresponds to a predetermined proportion of the region of interest of the window B _P.

For example, and in particular in the case of a peak P which is not located near other radiation peaks in the spectrum S, the windows B _b and B _H are taken with a region of interest equal to half of the region of interest in window B _P.

Windows B _b and B _{H have been} determined determined at the end of step 400 in FIG. 4d. The hits detected on the window B _H correspond to background noise, which may include photons having undergone Compton scattering and having lost energy after their emission.

In a step 500 of the method 20, the background noise observed on the windows B _b and B _{h is used} to interpolate the noise on the window B _P containing the peak P. The principle explained above is used in relation in Figures 3a to 3c, to suppress measurement noise.

We calculate count rate sums respectively on the windows B _b , B _p and B _h . These sums correspond respectively to the integral of the count rate curve of Figure 4d on the sets [N3; NEITHER NOR ; N2] and [N2; N4].

The photon count rate attributable to the source linked to the radiation peak P on the window B _P is then determined to be equal to the total count rate on the window B _{P from} which the total count rate is subtracted from the window B _b and the total count rate in window B _H.

In this example, by adding the regions of interest of the windows B _b and B _H , a region of interest of the band B _P is obtained.

Thus, the count rate attributable to the source obtained at the end of step 500 corresponds to the integral on the window B _P from which the average of the background noise is subtracted on the high energy windows B _H and low energy B _P.

The result thus obtained can be advantageously used in a subsequent step 600 to calculate the activity of the source to be identified and which is linked to the radiation peak P. In fact, the activity of the source is a function of the rate of counting on a source emission window, and other parameters such as the distribution of radionuclides in the field, the intrinsic efficiency of the detector, information on the topography of the field ...

The photon count rate attributable to the source to be identified and the activity of said source are for example linked according to the following mathematical expression:

[0116] Α = Ν / (Ι * ε * Τ),

With the activity of the radionuclide in becquerel (or in becquerel per unit of area, or in becquerel per unit of mass), N the number of strokes directly attributable to the radionuclide determined at the end of the process described above , I the emission probability of the energy line associated with the radionuclide in%, T the measurement acquisition time in seconds (600 seconds in the above example) and ε the intrinsic efficiency of the detector at a level d energy corresponding to the radionuclide line.

It is possible to carry out treatments subsequent to the calculation of this activity to determine a location of a radioactive source in the field. Predetermined models for calculating the probability of the presence of a source at a point, such as those of patent application FR 2 991 781 in the name of the Applicant, make it possible to obtain probabilities of angle of incidence of a photon detected at a given point by a non-collimated detector.

FIGS. 5a and 5b show a radiation spectrum S ’obtained at a point, to which an alternative embodiment of the method for determining the share of count rate attributable to a source is applied. The same references are used in these figures and in Figures 3b for similar elements.

On the spectrum S ', we note the presence of two peaks of radiation PI and P2 close together. The two peaks correspond to two energy lines separated from each other by only 36 keV. The application of steps 200 and 300 of method 20 gives, in the case of a window B _P of region of interest equal to the width at half height for the peak P ₂ , an energy window in which is also present the peak P _b

Low and high energy windows which would be defined as above, with a region of interest equal to half the region of interest of the window B _P , are likely to give a counting rate of the window of low energy B _b very close to the average count rate on window B _P. By subtracting the count rate from window B _b , we risk removing too large a proportion and underestimating the count rate attributable to the sources to be identified.

Thus, in this case with two peaks, at step 400, level N ₄ is determined, as shown in FIG. 4a, so that the high energy window B _H is of the same region of interest as window B _P. It is not necessary to determine a low energy window. Then, in step 500, to obtain the count rate attributable to the source for the peaks P] and P ₂ , the sum of count rate on window B _P is subtracted from the sum of count rate on window B _H , as shown in Figure 4b.

This embodiment has, for the case of two close radiation peaks, the advantage of allowing an estimation of counting rate which does not overestimate the background noise. This alternative embodiment can also be used for the case of three close radiation peaks, or for more than three peaks.

Claims (1)

[Claim 1] [Claim 2] [Claim 3] [Claim 4] [Claim 5] Claims Method for determining a share of photon count rate attributable to a photon source, from a radiation spectrum comprising a photon count rate associated with each level of a plurality of energy levels, the process comprising the steps of: - detection (200) of a local maximum count rate, said maximum constituting a peak (P) of radiation, determination (300) of a first energy level (Ni) and of a second energy level (N ₂ ), these two levels delimiting a first energy window comprising the peak, the method being characterized in that that it further includes the following steps: - determination (400) of a high level (N4) greater than the second level, defining with the second level a high energy window, the high level being determined as a function of a region of interest (ROI) of the first window energy, and preferably determination of a low level (N3) lower than the first level, delimiting with the first level a low energy window, the low level then being determined as a function of a region of interest ( ROI) of the first energy window, - calculation (500) of a part of the total counting rate of photons in the first energy window which is attributable to a source of photons emitting according to the radiation peak, by subtracting from said total counting rate the total counting rate in the high energy window, and possibly the total count rate in the low energy window. The method of claim 1, wherein the region of interest of the first energy window is defined using the width at half-height of the detected radiation peak. Method according to either of Claims 1 and 2, in which the region of interest of the high energy window is equal to the region of interest of the low energy window. The method of claim 3, wherein the regions of interest of the high and low energy windows are both equal to half the region of interest of the first energy window. Method according to one of claims 1 to 4, in which to determine it13 mination of the share of the total count rate attributable to the source, the total count rates are subtracted from the total count rate in the high energy window and in the low energy window. [Claim 6] Method according to one of claims 1 to 3, in which during the detection step, two radiation peaks associated with two distinct energy levels are detected in the spectrum, the first energy window comprising the two said peaks . [Claim 7] The method of claim 6, wherein the region of interest of the high energy window is equal to the region of interest of the first energy window. [Claim 8] Method according to either of Claims 6 and 7, in which for the determination of the share of the count rate attributable to the source, only the total count rate is subtracted from the high energy window. [Claim 9] Method according to one of claims 1 to 8, comprising a subsequent step of determining (600) an activity of a photon source which would be located at the geographical coordinates of a measurement point. [Claim 10] Method according to one of claims 1 to 9, further comprising a preliminary step of acquiring (100) the radiation spectrum at a measurement point. [Claim 11] A photon source identification system, the system comprising: a radiation detector (11) configured to measure a radiation spectrum, a processing unit (12) and a memory (13) associated with the processing unit, the detector being configured to transmit radiation spectrum measurements to the processing unit, the processing unit being configured to determine, from an acquired radiation spectrum, a share of photon count rate attributable to a source of photons by implementing the method according to one of claims 1 to 10. [Claim 12] System according to claim 11, the system being placed on a transport vector (10), the system being adapted to be moved over the surface of a ground, the radiation detector (11) being adapted to acquire spectra of radiation at a plurality of points on the surface. [Claim 13] The system of claim 11 or 12, further comprising a detector support structure, and a propulsion unit adapted to suspend the support structure above the surface of a [Claim 14] [Claim 15] field, the radiation detector (11) being adapted to acquire radiation spectra at a plurality of points on the surface. System according to one of Claims 11 to 13, in which the radiation detector is not collimated. Computer program product comprising code instructions for implementing the method according to any one of claims 1 to 10, when said instructions are executed by a processing unit.