GB2505897A

GB2505897A - Discriminating Radiation Pulses

Info

Publication number: GB2505897A
Application number: GB1216345.7A
Authority: GB
Inventors: William Barlow; Douglas Beverley Stevenson King; Aristide Eduard Ferdinand Mooyaart
Original assignee: BAE Systems PLC
Current assignee: BAE Systems PLC
Priority date: 2012-09-13
Filing date: 2012-09-13
Publication date: 2014-03-19
Also published as: GB2519726A; GB201216345D0; GB2519726B; WO2014041338A1; GB201503943D0

Abstract

Discriminating between different types of radiation; in particular gamma and beta. Template pulses are used for the types of radiation to be discriminated. Correlations (125, 136) are determined (s16, s36) between the shapes of the received pulses and the template pulses. The discriminating may further comprise normalizing the pulses, for example to the same height by controlling gain. Use of a neuronam process to determine correlations, plus Vernam pseudo-random binary key pattern, are referred to.

Description

DISCRIMINATING PULSE TYPES

FIELD OF THE INVENTION

The present invention relates to methods and systems for detecting, identifying, or discriminating between, types of pulses. The present invention relates in particular, but not exclusively, to detecting, identifying, or discriminating between, pulses induced by neutron radiation and pulses induced by gamma radiation in a radiation detection system.

BACKGROUND

Typical pulses output by radiation detection systems in response to neutron radiation 2 and in response to gamma radiation 4 respectively are shown by way of example in Figure 1 (the polarity of the pulse is irrelevant -the pulses may be represented as positive or negative going waveforms). The inset expanded plot on the right hand side of Figure 1 shows a magnified view of the difference in the slow components.

Conventionally, in order to discriminate between the types of pulses such as those discussed above, the rising slope and falling slope of the pulses are measured and compared. However, this is often problematic, for example when discriminating between neutron and gamma response pulses in many types of radiation detector, e.g. liquid or solid scintillators with photomultiplier type opto-electronic conversion, since there may often only be subtle differences in the rise and fall times.

SUMMARY OF THE INVENTION

The present inventors have realised it would be desirable to provide methods and systems for identifying types of pulses that have the potential to alleviate or remove disadvantages encountered by conventional methods when, for example, there are only subtle differences between the rise and fall times of different types of pulses that are to be discriminated between.

The present inventors have further realised it would be desirable for such methods and systems to take into account the shapes of the pulses being processed, in particular the whole extent of the shapes or substantially the whole extent of the shapes, as opposed to, for example, rise and/or fall times.

s In a first aspect, the present invention provides a method of discriminating between different types of pulses output by a radiation detection system; the method comprising performing the discriminating based on the different respective shapes of the different types of pulses.

In a further aspect, the present invention provides a system for discriminating between different types of pulses output by a radiation detection system; the system comprising means for performing the discriminating based on the different respective shapes of the different types of pulses.

In any of the above aspects, the following may apply.

The performing the discriminating may comprise: receiving the output pulses; generating respective template pulses for the types of pulses to be discriminated; and determining correlations between the shapes of the received pulses and the template pulses; differing levels of correlation determined for the respective different types of pulses providing the discrimination between the types of pulses.

The performing the discriminating may further comprise detecting the presence of the received output pulses and performing normalising so that the detected pulses are normalised to the template pulses and/or the template pulses are normalised to the detected pulses; and determining the correlations may comprise determining correlations between the normalised shapes of the received pulses and the template pulses and/or between the shapes of the received pulses and the normalised template pulses.

At least one of the template pulse types may be generated from a captured pulse of its type.

Performing the discriminating may further comprise performing Gray coding on the received pulses and as pad of the process of generating the template pulses.

Performing the discriminating may further comprise employing a pseudo-random binary key pattern.

The pseudo-random binary key pattern may be a Vernam encoding key pattern.

Determining correlation may comprise using a neuroram process.

A measure of a gain value employed in the normalising may be used to determine a measure of the energy of one or more of the detected pulses.

Performing the normalising may comprise the template pulses being normalised to the detected pulses; and the spectroscopic component of the detected pulses may be substantially retained.

One of the types of pulses being discriminated between may be pulses output by a radiation detection system in response to neutron radiation.

One of the types of pulses being discriminated between may be pulses output by a radiation detection system in response to gamma radiation.

In a further aspect, the present invention provides a program or plurality of programs arranged such that when executed by a computer system or one or more processors it/they cause the computer system or the one or more processors to operate in accordance with the method of any of the above aspects.

Ina further aspect, the present invention provides a machine readable storage medium storing a program or at least one of the plurality of programs of the above aspect.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows typical pulse shapes output by radiation detection systems in response to neutron radiation and in response to gamma radiation 4 respectively; Figure 2 is a block diagram of a system for discriminating between pulses output by a radiation detection system in response to neutron radiation and in response to gamma radiation; Figure 3 is a flowchart showing certain steps of a method of discriminating between neutron pulses and gamma pulses, as performed by the discrimination system of Figure 2; Figure 4 is a block diagram of a further discrimination system for discriminating between neutron pulses and gamma pulses output by a radiation detection system; Figure 5 is a flowchart showing certain steps of a method of discriminating between neutron pulses and gamma pulses, as performed by the discrimination system of Figure 4; Figure 6 shows an example of a simulated "idealised" neutron reference pulse generated by electronic circuitry; Figure 7 shows a generic template waveform of a neutron pulse produced from the reference waveform of Figure 6; Figure 8 shows an example of a simulated idealised" gamma reference pulse generated by electronic circuitry; Figure 9 shows a generic template waveform of a gamma pulse produced from the reference waveform of Figure 8; Figure 10 is a block diagram of a simplified from of an electronic circuit for providing reference neutron and gamma pulses; Figure 11 shows an example of a simulated test waveform with generated test neutron pulses and test gamma pulses; Figure 12 shows an example of a result obtained for neutron correlation response; Figure 13 shows an example of a result obtained for gamma correlation response; Figure 14 shows an example of a refined correlation response obtained when a further sum and thresholding process is applied to the neutron correlation response of Figure 12; and Figure 15 shows an example of a refined correlation response obtained when a further sum and thresholding process is applied to the gamma correlation response of Figure 13.

DETAILED DESCRIPTION

Figure 2 is a block diagram of an embodiment of a system 1 for discriminating between pulses output by a radiation detection system in response to neutron radiation 2 and in response to gamma radiation 4. For convenience, the system will be referred to hereinafter as the discrimination system 1 and the pulses will be referred to hereinafter as "neutron pulses 2" and "gamma pulses 4" respectively.

In this embodiment, the discrimination system 1 comprises the following modules: an analogue to digital (ND) converter 10, a delay module 12, a thresholder and peak detector 14, a normaliser 16, a local oscillator 18, a neutron pulse template generator 20, a gamma pulse template generator 22, and a correlator 24.

The analogue to digital (ND) converter 10 is arranged to receive input signals containing pulses that are to be discriminated between. An output of the ND converter is coupled to an input of the delay module 12 and to an input of the thresholder and peak detector 14. An output of the delay module 12 and also an output of the thresholder and peak detector 14 are coupled to respective inputs of the normaliser 16. A further output of the thresholder and peak detector 14 are coupled to respective inputs of the local oscillator 18, the neutron pulse template generator 20 and the gamma pulse template generator 22. An output of the local oscillator 18 is coupled to respective inputs of the neutron pulse template generator 20 and the gamma pulse template generator 22. Outputs of the normaliser 16, the neutron pulse template generator 20 and the gamma pulse template generator 22 are coupled to respective inputs of the correlator 24. The correlator 24 is arranged to output signals indicating discrimination between neutron pulses and gamma pulses.

Figure 3 is a flowchart showing certain steps of an embodiment of a method of discriminating between neutron pulses and gamma pulses, as performed by the discrimination system 1.

At step s2, an analogue waveform comprising a mixed neutron gamma signal, i.e. including both neutron pulses 2 and gamma pulses 4, is received as the input to the discrimination system 1. More particularly, the analogue waveform is received at the input to the AID converter 10.

At step s4, the analogue to digital (AID) converter 10 converts the signal to a digital signal. In this embodiment, this is performed such that the digital signal is Gray coded, as this is far more suitable for the later correlation process than natural binary. In other embodiments, other types of coding may be used, preferably other ones that, similar to Gray coding, reduce the Hamming distance between adjacent values in the coding scheme. Also, in other embodiments, keying using a pseudo-random binary pattern key may be used following the coding process (i.e. in this embodiment following the Gray coding), and this will advantageously tend to reduce local correlation maxima, as unwanted features tend to be swamped in the pseudo-random keying noise.

One particularly advantageous pseudo-random key for this purpose has been determined by the present inventors to be that used in a process known as Vernam encoding. Vernam encoding is described in US patent application no. US1 310,719, the contents of which are incorporated herein by reference. The digital signal is output to the thresholder and peak detector 14 and also to the delay module 12.

At step s6, the thresholder and peak detector 14 only allows signals above a predetermined threshold, i.e. noise floor, to be permitted through to its peak detector function, the peak detector function then determines the presence of any peaks, and also determines their respective amplitudes. For each detected peak, the presence of the peak, and its amplitude, are output to the normaliser 16. The presence of the peak (and optionally also its amplitude) is also output to the local oscillator 18, the neutron pulse template generator 20, and the gamma pulse template generator 22. Instead of the peak value as such, an appropriate gain setting may be sent to the normaliser 16.

The operation of the delay module 12 will now be described. The above described process being carried out by the thesholder and peak detector 14 requires a certain amount of time to perform. This time will be dependent upon specific characteristics of the detector, but will be known for the particular detector implementation. At step s8, the delay module 12 introduces a corresponding delay into the digital signal it has received from the AID converter 10, and outputs the delayed signal to the normaliser 16. In this embodiment the delay module is implemented in the form of a first-in-first-out shift register.

However, in other embodiments any other suitable implementation may be used. (In embodiments where the incoming signal is processed in analogue form rather than converting to digital, an analogue delay line can be used, for example one based on inductor-capacitor-resistor networks or one based on surface acoustic wave delay lines, e.g. using a lithium niobate substrate. Where appropriate or required, the incoming baseband signal may be modulated onto a carrier at a suitable frequency.) In this embodiment, at step slO, the normaliser 16 normalises the delayed signal being fed to it by the delay module 12 such that the peaks of all the detected pulses notified to it by the thresholding and peak detector 14 are made the same height as each other. In this embodiment, a multiplier is used, and the normaliser 16 makes use of the amplitude values received from the thresholding and peak detector 14. The nornialised signal is output to the correlator 24.

At step sl 2, in response to receiving the indication of the presence of the peak, the neutron pulse template generator 20 generates a template or reference waveform for a neutron pulse, and outputs this to the correlator 24.

Any suitable approach for generating the template may be used. In this embodiment, a template waveform of the neutron pulse has been pre-stored in a non-volatile memory of the neutron pulse template generator 20, and generating and outputting the template waveform comprises retrieving the stored shape from the memory and outputting it to the correlator 24. In this embodiment, the timing of this process is triggered by and synchronised to the timing of the detected input pulses by the local oscillator 18 which outputs suitable clock signals to the neutron pulse template generator 20 in response to receiving the indication of the peak from the thresholder and peak detector 14.

Returning to discussion of the pre-stored template waveform of the neutron pulse, this may be provided in any suitable way. In this embodiment, the shape of an actual neutron pulse as previously detected by the radiation system, or a shape "averaged" from plural detected actual neutron pulses as previously detected by the radiation system, for example in a setting up process, is used. See later below for further discussion of this shape "averaged" aspect.

In this embodiment the gamma pulse template generator 22 operates in a corresponding fashion to the neutron pulse template generator 20, as will now be described, although it is noted that in other embodiments the neutron pulse template generator 20 and the gamma pulse template generator 22 may operate in substantially different ways to each other.

In corresponding fashion, at step s14, in response to receiving the indication of the presence of the peak, the gamma pulse template generator 22 generates a template or reference waveform for a gamma pulse, and outputs this to the correlator 24. In response to receiving the indication of the presence of the peak, the gamma pulse template generator 22 generates a template or reference waveform for a gamma pulse, and outputs this to the correlator 24.

Any suitable approach for generating the template may be used. In this embodiment, a template waveform of the gamma pulse has been pre-stored in a non-volatile memory of the gamma pulse template generator 22, and generating and outputting the template waveform comprises retrieving the stored shape from the memory and outputting it to the correlator 24. In this embodiment, the timing of this process is triggered by and synchronised to the timing of the detected input pulses by the local oscillator 18 which outputs suitable clock signals to the gamma pulse template generator 22 in response to receiving the indication of the peak from the thresholder and peak detector 14.

Returning to discussion of the pre-stored template waveform of the gamma pulse, this may be provided in any suitable way. In this embodiment, the shape of an actual gamma pulse as previously detected by the radiation system, or a shape averaged from plural detected actual neutron pulses as previously detected by the radiation system, for example in a setting up process, is used.

In this embodiment, at step sl 6, the correlator 24 determines the level of correlation between detected and normalised input pulses received from the normaliser 16 and the template waveforms for a neutron pulse and the template waveforms for a gamma pulse as received from the neutron pulse template generator 20 and the gamma pulse template generator 22 respectively. Any suitable shape correlation process may be employed. In this embodiment, an approach called "neuroram" is used. Neuroram is described, for example, in International Patent publication number WO 99/33019, which is incorporated herein by reference. It is noted that the description of the neuroram correlation process given in WO 99/33019 includes the functionality, or equivalent functionality, already described above for the local oscillator 18, the neutron pulse template generator 20 and the gamma pulse template generator 22 in this embodiment. In overview, the neuroram process uses a neural pattern matcher that is made up of an array of first sum and threshold SAT1 devices each of which receives a number of inputs and a threshold value, and fires an output if the number of inputs exceeds the threshold value. The outputs of the array of the SAIl devices may be considered as a 2D image or generic template against which new data supplied into registers making up a data plane is correlated at a correlation plane of EXOR gates. The outputs of the EXOR gates themselves may be summed and thresholded by a second sum and threshold device to provide a neural output 1' or 0' indicating match or no match. The matcher may therefore behave as a neural auto-associative memory which continually adapts to the input data to recognise data of a particular specified class.

In those embodiments where Vernam encoding is employed, the generic template waveform is also keyed with the known pseudo-random binary pattern.

In this embodiment, the output from the correlator 24 accordingly indicates which of the detected input pulses are neutron pulses and which are gamma pulses. In most cases, an input pulse will have been clearly identified in the correlator 24 as a specific one of the types and clearly not the other type. In other borderline cases, where the correlator 24 would in isolation be satisfied that an input pulse correlated sufficiently to both a neutron pulse and a gamma pulse, final discrimination is achieved by a second sum and threshold, SAT2.

For example, this could be set to detect pulses that have a correlation score of 60%, with other pulses not being counted.

In this embodiment, as described above, the amplitude of the peaks, or a gain setting, is output to the normaliser 16 that then normalises the input signal.

Hence the peaks in the input signal are matched amplitude wise to the templates used later in the correlation. One advantage of this approach to matching the amplitude of the peaks to the amplitude of the templates is that the determined gain value may be used as a measure of the energy of the original peak, which value is of interest in many applications.

A further embodiment will now be described in which a further advantage is achieved whilst still being able, if desired, to use the gain value as a measure of the energy of the original peak. In overview, instead of the arrangement in the earlier described embodiment in which the input signal is normalised by the normaliser 16, in this further embodiment the template waveforms are normalised by normaliser modules comprised by the neutron pulse template generator 20 and the gamma pulse template generator 22. One advantage of this approach is that some or all of the spectroscopic component in the input signal is retained.

Figure 4 is a block diagram of this further embodiment of a discrimination system 101 for discriminating between neutron pulses 2 and gamma pulses 4 output by a radiation detection system.

In this further embodiment, the discrimination system 101 comprises the following modules: an analogue to digital (ND) converter 10, a delay module 12, a thresholder and peak detector 14, a local oscillator 18, a neutron pulse template generator 20, a gamma pulse template generator 22, and a correlator 24. The neutron pulse template generator 20 comprises a normaliser module 116, and the gamma pulse template generator 22 comprises a normaliser module 216.

The analogue to digital (ND) converter 10 is arranged to receive input signals containing pulses that are to be discriminated between. An output of the ND converter is coupled to an input of the delay module 12 and to an input of the thresholder and peak detector 14. An output of the thresholder and peak detector 14 is coupled to respective inputs of the local oscillator 18, the neutron pulse template generator 20, and the gamma pulse template generator 22. An output of the local oscillator 18 is coupled to respective inputs of the neutron pulse template generator 20 and the gamma pulse template generator 22.

Outputs of the delay module 12, the neutron pulse template generator 20, and the gamma pulse template generator 22 are coupled to respective inputs of the correlator 24. The correlator 24 is arranged to output signals indicating discrimination between neutron pulses and gamma pulses.

Figure 5 is a flowchart showing certain steps of a further embodiment of a method of discriminating between neutron pulses and gamma pulses, as performed by the discrimination system 101.

At step s22, an analogue waveform comprising a mixed neutron gamma signal, i.e. including both neutron pulses 2 and gamma pulses 4, is received as the input to the discrimination system 101. More particularly, the analogue waveform is received at the input to the ND converter 10.

At step s24, the analogue to digital (ND) converter 10 converts the signal to a digital signal. In this embodiment, this is performed such that the digital signal is Gray coded, as this is far more suitable for the later correlation process than natural binary. In other embodiments, other types of coding may be used, preferably other ones that, similar to Gray coding, reduce the Hamming distance between adjacent values in the coding scheme. Also, in other embodiments, keying using a pseudo-random binary pattern key may be used following the coding process (i.e. in this embodiment following the Gray coding), and this will advantageously tend to reduce local correlation maxima, as unwanted features tend to be swamped in the pseudo-random keying noise.

At step s26, the thresholder and peak detector 14 only allows signals above a predetermined threshold, i.e. noise floor, to be permitted through to its peak detector function, the peak detector function then determines the presence of any peaks, and also determines their respective amplitudes. For each detected peak, the presence of the peak, and its amplitude, are output to the local oscillator 18, the neutron pulse template generator 20, and the gamma pulse template generator 22 (the amplitude information need not be sent to the local oscillator 18 if preferred, in which case a separate output of the thresholder and peak detector 14 would be used). Instead of the peak value as such, an appropriate gain setting may be sent to the neutron pulse template generator 20, and the gamma pulse template generator 22.

The operation of the delay module 12 will now be described. The above described process being carried out by the thesholder and peak detector 14 requires a certain amount of time to perform. This time will be dependent upon specific characteristics of the detector, but will be known for the particular detector implementation. At step s28, the delay module 12 introduces a corresponding delay into the digital signal it has received from the AID converter 10, and outputs the delayed signal to the correlator 24. In this embodiment the delay module is implemented in the form of a first-in-first-out shift register.

However, in other embodiments any other suitable implementation may be used. (In embodiments where the incoming signal is processed in analogue form rather than converting to digital, an analogue delay line can be used, for example one based on inductor-capacitor-resistor networks or one based on surface acoustic wave delay lines, e.g. using a lithium niobate substrate. Where appropriate or required, the incoming baseband signal may be modulated onto a carrier at a suitable frequency.) At step s32, in response to receiving the indication of the presence of the peak and its amplitude or required gain setting, the neutron pulse template generator 20 generates a normalised template or reference waveform for a neutron pulse, and outputs this to the correlator 24. Any suitable approach for generating the template may be used. In this embodiment, a template waveform of the neutron pulse has been pre-stored in a non-volatile memory of the neutron pulse template generator 20, and generating and outputting the template waveform comprises retrieving the stored shape from the memory, the normaliser module 116 normalising the stored shape to the received amplitude/gain setting value (such that the peaks of the template waveform being generated is made the same height as the detected received peak it is being provided for), and the neutron pulse template generator 20 outputting the normalised form to the correlator 24. In this embodiment, the timing of this generation process is triggered by and synchronised to the timing of the detected input pulses by the local oscillator 18 which outputs suitable clock signals to the neutron pulse template generator 20 in response to receiving the indication of the peak from the thresholder and peak detector 14.

Returning to discussion of the pre-stored template waveform of the neutron pulse, this may be provided in any suitable way. In this embodiment, the shape of an actual neutron pulse as previously detected by the radiation system, or a shape "averaged" from plural detected actual neutron pulses as previously detected by the radiation system, for example in a setting up process, is used. See later below for further discussion of this shape "averaged" aspect. -14-

In this further embodiment the gamma pulse template generator 22 again operates in a corresponding fashion to the neutron pulse template generator 20, as will now be described, although it is noted that in other embodiments the neutron pulse template generator 20 and the gamma pulse template generator 22 may operate in substantially different ways to each other.

In corresponding fashion, at step s34, in response to receiving the indication of the presence of the peak and its amplitude or required gain setting, the gamma pulse template generator 22 generates a normalised template or reference waveform for a gamma pulse, and outputs this to the correlator 24.

Any suitable approach for generating the template may be used. In this embodiment, a template waveform of the gamma pulse has been pre-stored in a non-volatile memory of the gamma pulse template generator 22, and generating and outputting the template waveform comprises retrieving the stored shape from the memory, the normaliser module 216 normalising the stored shape to the received amplitude/gain setting value (such that the peaks of the template waveform being generated is made the same height as the detected received peak it is being provided for), and the gamma pulse template generator 22 outputting the normalised form to the correlator 24. In this embodiment, the timing of this generation process is triggered by and synchronised to the timing of the detected input pulses by the local oscillator 18 which outputs suitable clock signals to the gamma pulse template generator 22 in response to receiving the indication of the peak from the thresholder and peak detector 14.

In this embodiment, at step s36, the correlator 24 determines the level of correlation between detected input pulses received from the delay module 12 and the normalised template waveforms for a neutron pulse and the normalised template waveforms for a gamma pulse as received from the neutron pulse template generator 20 and the gamma pulse template generator 22 respectively. Any suitable shape correlation process may be employed. In this embodiment, an approach called "neuroram" is used, as described earlier above for the embodiment described with reference to Figures 2 and 3. The situation for borderline cases is also as described earlier above for the embodiment described with reference to Figures 2 and 3.

In the above embodiments, the shape of actual neutron pulses and gamma pulses as previously detected by the radiation system, or a shape "averaged" from plural detected actual neutron pulses and neutron pulses as previously detected by the radiation system, for example in a setting up process, are used as the template waveforms of the neutron and gamma pulses.

In other embodiments, other ways of generating template waveforms may be used. For example, a simulated or electronic-circuit generated reference pulse may be provided, from which a template waveform is then provided. One embodiment of this will now be described with reference to Figures 6 to 10.

Figure 6 shows an example of a simulated "idealised" neutron reference pulse (with arbitrary amplitude and time units, and not to scale, or at least not accurately scaled) generated by electronic circuitry. The amplitude axis is indicated by reference numeral 600 and the time axis is indicated by reference numeral 610. Using the general neuroram approach described in WO 99/33019, with appropriate minor routine adjustments by the skilled person if and when required, thirty two (this can be any appropriate number) waveforms with their own respective random noise profile are generated from the reference pulse of Figure 6, and Gray coding, and also "sum and thresholding", are performed, to arrive at one generic template waveform representing a form of an "average" of the shapes that will be produced in a noisy environment from a pure reference waveform starting point. This is somewhat analogous to the above mentioned "averaging" of the shape ofa numberof actual neutron orgamma pulseswhen they are used in the generation of the template waveform as described earlier above.

Figure 7 shows the generic template waveform of the neutron pulse produced in this example from the reference waveform of Figure 6. The amplitude axis is indicated by reference numeral 700 and the time axis is indicated by reference numeral 710. Again, the waveform is shown in terms of arbitrary amplitude and time units, and not to scale, or at least not accurately scaled.

Figure 8 shows an example of a simulated "idealised" gamma reterence pulse (with arbitrary amplitude and time units, and not to scale, or at least not accurately scaled) generated by electronic circuitry. The amplitude axis is indicated by reference numeral 800 and the time axis is indicated by reference numeral 810. Using the general neuroram approach described in WO 99/33019, with appropriate minor routine adjustments by the skilled person if and when required, thirty two (this can be any appropriate number) waveforms with their own respective random noise profile are generated from the reference pulse of Figure 8, and Gray coding, and also "sum and thresholding", are performed, to arrive at one generic template waveform representing a form of an "average" of the shapes that will be produced in a noisy environment from a pure reference waveform starting point.

Figure 9 shows the generic template waveform of the gamma pulse produced in this example from the reference waveform of Figure 8. The amplitude axis is indicated by reference numeral 900 and the time axis is indicated by reference numeral 910. Again, the waveform is shown in terms of arbitrary amplitude and time units, and not to scale, or at least not accurately scaled.

Any appropriate electronic circuit (or indeed computer simulation process) may be used to provide the initial reference pulses such as those shown in Figures 6 and 8. By way of example, a simplified from of a suitable electronic circuit for providing reference neutron and gamma pulses is shown in Figure 10. The portion of the circuit indicated by reference numeral 150 provides simulated neutron pulses and the portion of the circuit indicated by reference numeral 160 provides simulated gamma pulses. The circuits are fed by narrow rectangular pulses from a signal generator and yield waveforms representing neutron and gamma pulses.

s Results provided by running a MATLAB (trademark) simulation using waveforms of Figures 6 to 9 above for generation of the template waveforms will now be described with reference to Figures 11 to 15.

Figure 11 shows an example of a simulated test waveform (with arbitrary amplitude and time units, and not to scale, or at least not accurately scaled) with generated test neutron pulses 115 and test gamma pulses 116. The amplitude axis is indicated by reference numeral 111 and the time axis is indicated by reference numeral 112. A preamble and a postamble are not shown.

The waveform of Figure 11 was applied as a test signal to a simulation of the discriminating subsystem 1 including neuroram correlation as described and referenced earlier above.

Figure 12 shows an example of the result obtained for neutron correlation response, with arbitrary amplitude and time units, and not to scale, or at least not accurately scaled, although the time axis is kept approximately the same as that of Figure 11. The amplitude axis is indicated by reference numeral 121 and the time axis is indicated by reference numeral 122. As can be seen, the highest correlation peaks 125 are successfully obtained at times matching the times of the neutron pulses 115 in the test signal of Figure 11.

Figure 13 shows an example of the result obtained for gamma correlation response, with arbitrary amplitude and time units, and not to scale, or at least not accurately scaled, although the time axis is kept approximately the same as that of Figures 11 and 12. The amplitude axis is indicated by reference numeral 131 and the time axis is indicated by reference numeral 132. As can be seen, the highest correlation peaks 136 are successfully obtained at times matching the times of the gamma pulses 116 in the test signal of Figure 11.

An additional sum and thresholding process may be applied to the results of Figures 12 and 13, to further refine the correlation results. By use of suitable threshold levels, the response at the correlation peaks can be further discriminated from the remained of the correlation result. This further sum and thresholding step is described as "SAT2" in WO 99/3301 9. The outcome of such a further sum and thresholding process is shown in Figures 14 and 15.

Figure 14 shows an example of the refined correlation response obtained when the further sum and thresholding process is applied to the neutron correlation response of Figure 12. The response is represented with arbitrary amplitude and time units, and not to scale, or at least not accurately scaled, although the time axis is kept approximately the same as that of Figures 11 to 13. The amplitude axis is indicated by reference numeral 141 and the time axis is indicated by reference numeral 142. As can be seen, very clear correlation peaks 145 are successfully obtained at times matching the times of the neutron pulses 115 in the test signal of Figure 11, with no erroneous responses or noise passing through the additional thresholding stage.

Figure 15 shows an example of the refined correlation response obtained when the further sum and thresholding process is applied to the gamma correlation response of Figure 13. The response is represented with arbitrary amplitude and time units, and not to scale, or at least not accurately scaled, although the time axis is kept approximately the same as that of Figures 11 to 14. The amplitude axis is indicated by reference numeral 151 and the time axis is indicated by reference numeral 152. As can be seen, very clear correlation peaks 155 are successfully obtained at times matching the times of the gamma pulses 116 in the test signal of Figure 11, with no erroneous responses or noise passing through the additional thresholding stage.

Apparatus, including the discriminating systems 1 and 101, for implementing the above arrangement, and performing the method steps to be described later below, may be provided by configuring or adapting any suitable apparatus, including for example one or more computers or other processing apparatus or processors, and/or providing additional modules. The apparatus may comprise a computer, a network of computers, or one or more processors, for implementing instructions and using data, including instructions and data in the form of a computer program or plurality of computer programs stored in or on a machine readable storage medium such as computer memory, a computer disk, ROM, PROM etc., or any combination of these or other storage media.

It should be noted that certain of the process steps depicted in the flowcharts of Figures 3 and 5 and described above may be omitted or such process steps may be performed in differing order to that presented above and shown in the Figures. Furthermore, although all the process steps have, for convenience and ease of understanding, been depicted as discrete temporally-sequential steps, nevertheless some of the process steps may in fact be performed simultaneously or at least overlapping to some extent temporally.

In the above embodiments, discrimination is performed between neutron pulses and gamma pulses. However, this need not be the case, and in other embodiments discrimination may be performed between any two types of pulses output by radiation detection pulses in response to any types of radiation, for example neutron compared to a type other than gamma, gamma compared to a type other than neutron, or between two different types other than neutron or gamma.

Moreover, although in the above embodiments discrimination between two types of pulses is performed, in other embodiments the process may be used simply for identifying the type of a detected pulse, for example identifying neutron pulses alone, identifying gamma pulses alone, or identifying any other particular type pulse detected from radiation. In yet further embodiments, discrimination between more than two types of pulses may be performed, by the use of more than two respective template waveforms.

In the above embodiments a digitised shape correlation process is used, in particular using neuroram as discussed and referenced above, and alternatively based on neuroram as discussed and referenced above. However, in other embodiments other digital shape correlation processes may be used other than neuroram or ones based on neuroram. In this regard, it is noted that the received pulses and the generated template waveforms can be put into -20 -phase and made of approximately equal amplitudes if desired (as performed in the above embodiments). One simple example of an alternative form of digitised correlator that may be used is to use a correlator comprising an array of Exclusive-NOR gates. The correlation bit results may be counted using binary counters, with the greatest count representing the "winning" i.e. identified pulse, either neutron or gamma. If the counts were the same then the result would be deemed uncertain.

In yet further embodiments, an analogue shape correlation process may be used instead of a digital shape correlation process. In one example an analogue correlator may use a simple difference circuit (a subtractor) together with an integrator -this is similar to the action of a neuron. The integrator with the greatest value at the end of the specified correlation time window represents the "winning" i.e. identified pulse, either neutron or gamma. It could be possible that the integrators have the same or approximately the same value -in this case the result would be deemed uncertain. The system could be extended to separately count the neutron, gamma or "uncertain" pulses.

In the above embodiments a synchronised local oscillator is involved in the process of generating the template pulses. However, this is not essential, and in other embodiment this may be omitted or a different method of temporal synchronising may be used.

As mentioned earlier above, in the above described embodiments, the peaks in the input signal are matched amplitude wise to the templates used later in the correlation (or vice versa). One advantage of this approach is that the determined gain value may be used as a measure of the energy of the original peak, which value is of interest in many applications.

As mentioned earlier above, in some of the above described embodiments, a further advantage is achieved whilst still being able, if desired, to use the gain value as a measure of the energy of the original peak. In these embodiments the template waveforms are normalised to the received input peaks (as opposed to the received input peaks being normalised to the -21 -template waveforms). One advantage of this approach is that some or all of the spectroscopic component in the input signal is retained.

Claims

-22 -CLAIMS1. A method of discriminating between different types of pulses (2, 4) output by a radiation detection system; the method comprising: performing the discriminating based on the different respective shapes of the different types of pulses (2, 4).
2. A method according to claim 1, wherein performing the discriminating comprises: receiving (s2; s22) the output pulses (2, 4); generating (s12, s14; s32, s34) respective template pulses for the types of pulses to be discriminated; and determining (s16, s36) correlations (125, 136) between the shapes of the received pulses and the template pulses; differing levels of correlation determined for the respective different types of pulses providing the discrimination between the types of pulses.
3. A method according to claim 2, wherein performing the discriminating further comprises detecting (s6; s26) the presence of the received output pulses (2, 4) and performing normalising so that the detected pulses are normalised to the template pulses and/or the template pulses are normalised to the detected pulses; and determining (s16; s36) the correlations (125, 136) comprises determining correlations between the normalised shapes of the received pulses and the template pulses and/or between the shapes of the received pulses and the normalised template pulses.

-23 -
4. A method according to claim 2 or claim 3, wherein at least one of the template pulse types is generated from a captured pulse of its type.
5. A method according to any of claims 2 to 4, wherein performing the discriminating further comprises performing Gray coding (s4; s24) on the received pulses and as part of the process of generating the template pulses.
6. A method according to any of claims 2 to 5, wherein performing the discriminating further comprises employing a pseudo-random binary key pattern.
7. A method according to claim 6, wherein the pseudo-random binary key pattern is a Vernam encoding key pattern.
8. A method according to any of claims 2 to 7, wherein determining correlation comprises using a neuroram process.
9. A method according to any of claims 3 to 8, wherein a measure of a gain value employed in the normalising is used to determine a measure of the energy of one or more of the detected pulses.
10. A method according to any of claims 3 to 9, wherein performing the normalising comprises the template pulses being normalised to the detected pulses; and the spectroscopic component of the detected pulses is substantially retained.

-24 -
11. A method according to any of claims 1 to 10, wherein one of the types of pulses being discriminated between is pulses (2) output by a radiation detection system in response to neutron radiation.
12. A method according to any of claims 1 to 11, wherein one of the types of pulses being discriminated between is pulses (4) output by a radiation detection system in response to gamma radiation.
13. A system (1; 101) for discriminating between different types of pulses (2, 4) output by a radiation detection system; the system comprising: means for performing the discriminating based on the different respective shapes of the different types of pulses (2, 4).
14. A program or plurality of programs arranged such that when executed by a computer system or one or more processors it/they cause the computer system or the one or more processors to operate in accordance with the method of any of claims 1 to 12.
15. A machine readable storage medium storing a program or at least one of the plurality of programs according to claim 14.