CN109816740A - A Coincidence Processing Method of Flicker Pulse Event - Google Patents

Abstract

The present invention provides a method for coincident processing of scintillation pulse events, comprising step S1: acquiring experimental data of a PET system, the experimental data including a pair of gamma photons with the same energy and opposite directions generated after positron annihilation reaching two corresponding detectors The time information and the energy information deposited by the gamma photons in the detector, the time information is T _i1 and T _i2 respectively, and the energy information is E _i1 and E _i2 respectively; Step S2: Set the energy window and time window and obtain the coincidence events , the energy window is set to E, the time window is set to T, and the gamma photons satisfying the following conditions are recorded as a pair of true coincidence events: ∣T _i1 -T _i2 ∣≤T, E _i1 ≤E, and E _i2 ≤E ; Step S3: carry out weighting processing on the true coincidence events, and two formulas are used for the enhancement processing; Step S4: carry out image reconstruction, and obtain image results; Step S5: analyze the image results. The invention uses the energy information to distribute the weights of the coincident events, improves the quality of the reconstructed image, and has strong adaptability.

Description

A Coincidence Processing Method of Flicker Pulse Event

technical field

The invention relates to a scintillation pulse processing method in the field of positron emission computed tomography, and more particularly to a coincidence processing method for scintillation pulse events.

Background technique

PET (Positron Emission Tomography, positron emission tomography, hereinafter referred to as PET) is a non-invasive nuclear medicine imaging method. After the radioactive tracer is injected into the human body or animal, the tracer will show different concentration distributions in different parts of the human body or animal body according to the metabolic level in different parts of the human body or animal body. The β+ decay of radioactive substances in the human body or animal produces positrons, and the positrons annihilate with electrons in the human body or animals to produce a pair of gamma photons with the same energy but opposite directions. Therefore, the gamma photons are measured by an external detection device. With the time information, energy information and position information arriving at the detection device, the distribution level of the radioactive tracer in the human body or animal body can be calculated, and the image can be reconstructed and displayed. PET is a functional metabolic molecular imaging device, which plays a very important role in the research, diagnosis and treatment of oncology and some diseases related to the nervous system.

After the positron annihilates in the human body, a pair of gamma photons with opposite directions of motion will be generated, and their energy values are both 511KeV. The gamma photons can be converted into scintillation pulse signals through the photoelectric conversion device and subsequent processing circuits. The time information, energy information and position information of gamma photons reaching the detector module can be measured by the scintillation pulse signal. Therefore, the coincidence of scintillation pulse events is a very important step in the PET image reconstruction process. The reconstructed image quality in the same situation is greatly improved. At present, the coincidence method of scintillation pulse events is mainly to establish a time window and an energy window to determine whether two scintillation pulse events are true coincidence events. When the absolute value is within the time window and the energy values of the two scintillation pulses are both within the energy window, the two scintillation pulses can be regarded as a pair of true coincidence events in the event coincidence project. The corresponding response lines of the pair of coincident events in the PET system can be obtained from the position information of the scintillation pulse. In the existing image reconstruction algorithm, due to the large number of crystal strips, when a pair of scintillation pulses is measured between crystal strip A and crystal strip B, the response lines corresponding to the two crystal strips are in the back-projection step of the iterative reconstruction. Backprojection is performed when it is a coincidence event (+1 when expressed as a count). However, in this reconstruction method, for each pair of scintillation pulses, each gamma photon may be scattered before hitting the detector or encounter the measurement error of the detector, and the energy value of the measured gamma photon is not It must be 511KeV, but in back-projection, these coincident event pairs are still regarded as true coincident events for back-projection. For example, when two pairs of gamma photons are measured, the time information measured by the first pair of gamma photon detectors is Ta1. , Ta2, the energy information is Ea1, Ea2 respectively; the time information measured by the second pair of gamma photon detectors are Tb1, Tb2, respectively, and the energy information is Eb1, Eb2, respectively, the time window of the PET system is set to T, the energy The window is set to E. In this method, when |Ta1-Ta2|≤T, Ea1≤E and Ea2≤E, the first pair of gamma photons is considered to be a pair of true coincidence events, and the same applies to judgment The second pair of gamma photons. However, Ea1 in actual measurement is not necessarily equal to Eb1, and Ea2 is not necessarily equal to Eb2. Therefore, it is unreasonable to treat all weights of coincident events as equal quantities.

There is also a weighted processing method based on time information in the prior art, namely TOF (time of flight, time of flight) reconstruction method, which is a reconstruction method that utilizes the time information of gamma photons arriving at the detector for weighting . However, because this reconstruction method has high requirements on the performance of the detector, it is mostly only used to reconstruct the simulation data.

To sum up, the coincidence method of scintillation pulse events using time window and energy window in the prior art has certain defects in theory because it is considered that each pair of coincident events has equal weights in back projection; Although the method of weighting time information is feasible, it has too high requirements for the time measurement performance of the detector, and cannot be widely used in practice. Therefore, in order to improve the quality of image reconstruction in the PET system, it is necessary to find a more perfect and practical method for the coincidence of scintillation pulse events.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method for coincidence processing of scintillation pulse events, so as to solve the problem of errors in the coincidence processing of scintillation pulse events in the prior art and inconvenience for practical application.

In order to solve the above-mentioned technical problems, the technical solution of the present invention is to provide a coincidence processing method of a scintillation pulse event, a coincidence processing method of a scintillation pulse event, and the coincidence processing method includes the following steps:

Step S1: Acquire experimental data of the PET system, wherein the experimental data includes the time information of a pair of gamma photons with the same energy and opposite directions generated after positron annihilation reaching the corresponding two detectors and the time information of the gamma photons at the corresponding detectors. The energy information deposited in the detector, the time information is respectively T _i1 and T _i2 , and the energy information is E _i1 and E _i2 respectively;

Step S2: Set an energy window and a time window of the PET system and obtain a coincident event, wherein the energy window is set to E, the time window is set to T, and the gamma photon record that satisfies the following conditions: For a pair of true matching events:

∣T _i1 -T _i2 ∣≤T, E _i1 ≤E, and E _i2 ≤E;

Step S3: Perform weighting processing on the true coincidence events, wherein the strengthening processing adopts the following two formulas:

The first:

The second:

where pi represents the projection value on the _i -th response line, i is a natural number, σ and a are variable parameters related to the PET system, imaging prosthesis and activity, respectively;

Step S4: performing image reconstruction to obtain image results;

Step S5: analyze the image results.

In the step S1, the experimental data further includes position information of gamma photons, and the position information is obtained by numbering the scintillation crystal strips of the PET system.

In the step S2, the setting range of the energy window T is 150keV-650KeV, and the energy resolution is set to 20%.

According to an embodiment of the present invention, in the step S3, the value range of σ is 0.1-1000.

According to an embodiment of the present invention, the value range of σ is 1.5-100.

According to an embodiment of the present invention, in the step S3, the value range of a is 0.01-100.

According to an embodiment of the present invention, the value of a ranges from 1 to 50.

According to an embodiment of the present invention, the value of σ is 2, and the value of a is 0.5.

In the step S4, the image reconstruction method is the maximum likelihood-expectation maximum algorithm, and the formula used is:

in, Represents the value of the jth pixel after n+1 iterations, j is a natural number, and a _ij represents the proportion of the jth pixel on the i-th response line.

In the step S5, the contrast parameter used for analyzing the image results is a restoration degree contrast coefficient, and the restoration degree contrast coefficient includes a hot area restoration degree contrast coefficient and a cold area restoration degree contrast coefficient, wherein the hot area restoration degree contrast coefficient The regional recovery degree contrast coefficient is:

Among them, CH _,j represents the average count of the hot area in the reconstructed image, C _B,j represents the average count of the background in the reconstructed image, a _H represents the activity value of the hot area during the simulation, and a _B represents the activity of the background during the simulation. value.

The cold zone recovery degree contrast coefficient is:

where C _C,j is the average count of cold regions in the reconstructed image, and C _B,j is the average count of the background in the reconstructed image.

In the coincidence processing method of the scintillation pulse event provided by the present invention, in step S3, a unique weighting processing method is adopted, each pair of response lines is weighted according to the energy value of the gamma photon, and different weights are given to each pair of response lines, It can maximize the utilization of the energy information obtained in the data acquisition process, and use the energy information to assign the weights of the matching events, so as to improve the quality of the reconstructed image in the subsequent image reconstruction, especially the steps can be significantly improved. The recovery degree contrast coefficient of the hot area and the contrast coefficient of the cold area recovery degree in S5; the present invention can also select different weighting methods or different parameters according to the imaging characteristics of different living bodies to improve the image reconstruction quality in a more targeted manner. Strong sex. At the same time, the method for coherent processing of the scintillation pulse event provided by the present invention is simple to implement, and can be applied to PET systems of various structures.

Description of drawings

FIG. 1 is a schematic diagram of steps of a method for processing scintillation pulse events according to a preferred embodiment of the present invention.

Detailed ways

The present invention will be further described below with reference to specific embodiments. It should be understood that the following examples are only used to illustrate the present invention and not to limit the scope of the present invention.

The coincidence processing method of the scintillation pulse event provided by the present invention is used for data processing in the PET system, and mainly comprises the following steps:

Step S1 : obtaining experimental data, the experimental data is the time information of the gamma photons generated after the positron annihilation reaches the detector, the energy information of the gamma photons deposited in the detector, and the position information of the gamma photons, the corresponding two The time information of the gamma photons collected by the detector are respectively T _i1 and T _i2 , and the energy information is respectively E _i1 and E _i2 ;

Step S2: Set an energy window and a time window, obtain a coincident event, and set the energy window and time window in the PET system as E and T respectively, when the time information and energy information of the gamma photon meet the following conditions respectively:

∣T _i1 -T _i2 ∣≤T,

E _i1 ≤ E

E _i2 ≤ E;

Then it is determined that the gamma photons collected by the corresponding detector are a pair of true coincidence events;

Step S3: Perform weighting processing on the true coincidence events, and the weighting processing methods include two ways:

The first:

The second:

Among them, pi represents the projection value on the _i -th response line, i is a natural number, σ and a are variable parameters related to the PET system, imaging prosthesis and activity, and different parameters can be selected according to needs in the process of image reconstruction σ and a of ;

Step S4: Perform image reconstruction. The specific method of image reconstruction is Maximum Likelihood Expectation Maximization (ML-EM for short), and the formula used in the maximum likelihood-expectation maximization algorithm is:

in, Represents the value of the jth pixel after n+1 iterations, j is a natural number, a _ij represents the proportion of the jth pixel on the ith response line, and p _i represents the projection value on the ith response line;

The fifth step is the image result analysis. The contrast parameter used in the image result analysis is the restoration degree contrast coefficient, and the restoration degree contrast coefficient includes the hot zone restoration degree contrast coefficient and the cold zone restoration degree contrast coefficient, wherein the hot zone restoration degree contrast coefficient is :

Among them, CH _,j represents the average count of the hot area in the reconstructed image, C _B,j represents the average count of the background in the reconstructed image, a _H represents the activity value of the hot area during the simulation, and a _B represents the activity of the background during the simulation. value.

The cold zone recovery degree contrast coefficient is:

where C _C,j is the average count of cold regions in the reconstructed image, and C _B,j is the average count of the background in the reconstructed image.

More specifically, in the above step S1, the energy window range set when acquiring the experimental data is 150KeV-650KeV, the energy resolution is set to 20%, and the prosthesis used to acquire the experimental data is the National Electrical Manufacturers Association (National Electrical Appliance Manufacturers Association). For the IQ (image quality) prosthesis under the Manufacturers Association (NEMA) NU 2-2007 standard, the parameters set by the PET system are as follows: the detector radius is 371mm, the number of detector plates is 88, and each detector plate includes 1 *1*4 arrayed detector modules, each detector module includes 1*6*6 arrays of scintillation crystal strips, the total number of scintillation crystal strips is 12672, and the size of a single scintillation crystal strip is 4.25mm* 4.25mm*4.25mm, the number of pixels is 500*500*96; the position information for obtaining the experimental data is obtained by numbering the scintillation crystal strips.

In the above step S2, the acquisition of the coincident events and the data comparison may be performed by a processor suitable for data operation.

In the above step S3, the value range of σ is preferably 0.1 to 1000. The value range of σ is more preferably 1.5 to 100. The value range of a is preferably 0.01 to 100. The value range of a is more preferably 1-50. In the preferred embodiment of the present invention, the value of σ is 2, and the value of a is 0.5.

When the traditional PET system is intensified processing, the energy information is discarded immediately after the coincidence event processing. The coincidence processing method of the scintillation pulse event provided by the present invention adopts a unique weighting processing method in step S3, according to the energy of gamma photons. The value of each pair of response lines is weighted, and each pair of response lines is given different weights, which can maximize the utilization of the energy information obtained in the data acquisition process, and use the energy information to assign the weights that match the events, so that it can be used later. It plays a role in improving the quality of the reconstructed image in the reconstructed image, and especially can significantly improve the contrast coefficient of recovery in the hot area and the contrast coefficient in the cold area in step S5; the present invention can also select different images according to the imaging characteristics of different living bodies. The weighting method or different parameters can improve the quality of image reconstruction in a more targeted manner, and the adaptability is strong. At the same time, the method for coherent processing of the scintillation pulse event provided by the present invention is simple to implement, and can be applied to PET systems of various structures.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention. Various changes can be made to the above-mentioned embodiments of the present invention. That is, all simple and equivalent changes and modifications made according to the claims and descriptions of the present invention fall into the protection scope of the claims of the present invention. What is not described in detail in the present invention is conventional technical content.

Claims (10)

1. A coincidence processing method of a scintillation pulse event, characterized in that the coincidence processing method comprises the steps of:

step S1: acquiring experimental data of a PET system, wherein the experimental data comprises time information of arrival of a pair of gamma photons with same energy and opposite directions generated after positron annihilation at two corresponding detectors and energy information of deposition of the gamma photons in the detectors, and the time information is T_i1And T_i2The energy information is respectively E_i1And E_i2；

Step S2: setting an energy window and a time window of the PET system and acquiring coincidence events, wherein the energy window is set as E, the time window is set as T, and the gamma photons meeting the following conditions are marked as a pair of true coincidence events:

∣T_i1－T_i2∣≤T，E_i1e is not more than E, and E_i2≤E；

Step S3: weighting the true coincidence events, wherein the strengthening processing adopts the following two formulas:

the first method comprises the following steps:

and the second method comprises the following steps:

wherein p is_iRepresenting a projection value on the ith response line, i is a natural number, and sigma and a are variable parameters related to the PET system, the imaging prosthesis and the activity respectively;

step S4: carrying out image reconstruction to obtain an image result;

step S5: and analyzing the image result.

2. The coincidence processing method of scintillation pulse events of claim 1, wherein in said step S1, said experimental data further comprises position information of gamma photons, said position information being obtained by numbering scintillation crystal strips of said PET system.

3. The coincidence processing method of scintillation pulse events of claim 1, wherein in step S2, the energy window T is set to a range of 150keV to 650keV and the energy resolution is set to 20%.

4. The coincidence processing method of the scintillation pulse event according to claim 1, characterized in that, in step S3, σ has a value in the range of 0.1-1000.

5. The coincidence processing method of scintillation pulse events of claim 4, wherein σ ranges from 1.5 to 100.

6. The coincidence processing method of the scintillation pulse event according to claim 1, wherein in step S3, a has a value in the range of 0.01 to 100.

7. The coincidence processing method of scintillation pulse events of claim 6, wherein a ranges from 1 to 50.

8. The coincidence processing method of scintillation pulse events of claim 1, wherein σ has a value of 2 and a has a value of 0.5.

9. The coincidence processing method of the scintillation pulse event of claim 1, wherein in said step S4, the method of image reconstruction is a maximum likelihood-expectation maximization algorithm, and the formula adopted is:

wherein,represents the value of j pixel after n +1 times of iteration, j is a natural number, a_ijIndicating the proportion of the j pixel on the ith response line.

10. The coincidence processing method of the scintillation pulse event of claim 1, wherein in step S5, the image result is analyzed by using a recovery contrast coefficient as a contrast parameter, the recovery contrast coefficient includes a hotspot recovery contrast coefficient and a cold-area recovery contrast coefficient, wherein the hotspot recovery contrast coefficient is:

wherein, C_H,jRepresenting the average count of hot areas in the reconstructed image, C_B,jRepresenting the average count of the background in the reconstructed image, a_HRepresenting the activity value of the hot zone in simulation, a_BThe activity value of the background at the time of simulation is represented.

The cold zone recovery contrast ratio is:

wherein, C_C,jRepresenting the average count of cold regions in the reconstructed image, C_B,jRepresenting the average count of the background in the reconstructed image.