CN106491153B

CN106491153B - A kind of PET scatter correction methods, PET imaging methods and PET imaging systems

Info

Publication number: CN106491153B
Application number: CN201611244336.8A
Authority: CN
Inventors: 朱闻韬; 李弘棣
Original assignee: Shanghai United Imaging Healthcare Co Ltd
Current assignee: Shanghai United Imaging Healthcare Co Ltd
Priority date: 2016-12-29
Filing date: 2016-12-29
Publication date: 2017-10-27
Anticipated expiration: 2036-12-29
Also published as: CN107137102A; CN107137102B; CN106491153A

Abstract

The application provides a kind of PET scatter correction methods, including：During PET scan is performed to person under inspection, the TOF PET datas of person under inspection target area are obtained, TOF PET datas include meeting event data and TOF data；According to the PET image for meeting event data acquisition target area；Based on PET image and TOF PET datas, the initial scatter estimation being imaged for TOF PET is obtained；Processing is corrected to initial scatter estimation, the scattering estimation of TOF PET imagings is obtained；Correction is scattered to TOF PET datas based on the scattering estimation that TOF PET are imaged.The application can improve the precision of scattering estimation, improve the quality of TOF PET images.Meanwhile, the application also proposes PET imaging methods and system.

Description

PET scattering correction method, PET imaging method and PET imaging system

Technical Field

The present application relates to the field of Positron Emission Tomography (PET), and in particular, to a Time of Flight (TOF) based scatter correction method, a PET imaging method, and a PET imaging system in a PET imaging process.

Background

Positron emission computed tomography is a noninvasive medical imaging technology for detecting the metabolic characteristics of human or animal organs and has the characteristics of high sensitivity, good accuracy and accurate positioning. In PET imaging, a radiopharmaceutical capable of emitting positrons is injected into a living body, the positrons and electrons in the living body are annihilated to generate two gamma photons with opposite directions and 511keV energy, a detector surrounding the detected living body detects the gamma photons and stores the information of the gamma photons in a coincidence counting mode, and if the recorded coincidence counting is enough, the original data can be reconstructed into a tomographic image of the concentration distribution of the radionuclide in the detected object through an image reconstruction algorithm.

Time-of-flight technology positron emission tomography (TOF-PET) scanners are an advanced functional imaging tool in nuclear medicine imaging, and the application prospect thereof has been highly valued by nuclear medicine imaging researchers and equipment manufacturers. TOF-PET has essentially the same composition as the imaging process and system of conventional PET, but the corresponding time measurement system is essentially different: in conventional PET, the function of the temporal measurement system is to determine whether true coincidence occurs within a coincidence time window in order to obtain projections of organs and tissue structures; in TOF-PET, the function of the time measurement system is to determine the position and intensity of the distribution of the radionuclide in a coincidence time window, and the time difference of two 511keV gamma rays generated by positron annihilation reaching a detector is utilized to locate the position of an annihilation event on a Line of Response (LOR) according to the light speed, so that the imaging quality of a PET scanner is improved, the dosage is reduced, and the scanning time is shortened.

However, in the latest TOF-PET scanners, although coincidence counting acquisition of all oblique lines of response improves the sensitivity of the system compared to the conventional 2D and extended 2D data acquisition modes, a large number of scatter coincidence counts are introduced at the same time, thereby reducing the resolution of the system and the quantitative accuracy of the imaging; also, in data acquisition using lead or tungsten baffles, photon scattering from the baffles can have a significant impact on the resolution and contrast of PET imaging. In view of this, in order to improve the accuracy of TOF-PET imaging, it is necessary to perform efficient scatter correction for TOF-PET systems.

Disclosure of Invention

It is an object of the present application to provide a scatter correction method for TOF-PET imaging that can improve scatter estimation accuracy.

In order to solve the above problem, according to an aspect of the present application, there is provided a TOF-PET scatter correction method including:

during a PET scan of a subject, acquiring TOF-PET data of a target region of the subject, the TOF-PET data including coincidence event data and TOF information;

acquiring a PET image of a target area according to the coincidence event data;

obtaining an initial scatter estimate for TOF-PET imaging based on the PET image and the TOF-PET data;

correcting the initial scattering estimation to obtain a scattering estimation of TOF-PET imaging;

scatter correcting the TOF-PET data based on the scatter estimate of the TOF-PET imaging.

Optionally, denoising the initial scatter estimate comprises: and performing high-frequency filtering processing on the initial scattering estimation to remove high-frequency components contained in the initial scattering estimation.

Optionally, performing a correction process on the initial scatter estimate to obtain a scatter estimate for TOF-PET imaging, including:

acquiring a TOF-PET scattering benchmark by using a single correction method based on flight time;

fitting the initial scattering estimation according to the TOF-PET scattering reference to obtain fitting parameters;

obtaining a scatter estimate for TOF-PET imaging based on the fitting parameters and the TOF scattered fiducial.

fitting the initial scattering estimation by using TOF scattering estimation mutually consistent with the PET image to obtain fitting parameters;

obtaining a TOF scatter estimate for TOF-PET imaging based on the fitting parameters and the initial scatter estimate.

According to another aspect of the present application, a PET imaging method is presented, comprising:

acquiring a PET image of a target area according to the coincidence event data;

scatter correcting the TOF-PET data based on a scatter estimate of the TOF-PET imaging; and reconstructing the scatter-corrected TOF-PET data to obtain a TOF-PET image of the target region.

obtaining a scatter estimate for TOF-PET imaging based on the fit parameters and the reference for TOF-PET scatter.

Optionally, the coincidence event data, the TOF information and/or the TOF-PET data are stored in a sinogram mode.

Optionally, the method further comprises:

acquiring random coincidence event data of a target region of a detected object, and randomly correcting the TOF-PET data according to the random coincidence event data;

or/and obtaining an anatomical image of a target region of a detected object, obtaining attenuation values corresponding to all voxels of the target region according to the anatomical image, and performing attenuation correction on the TOF-PET data according to the attenuation values.

In accordance with yet another aspect of the present application, a PET imaging system is presented, comprising a gantry, a detector ring, a pre-processor, a simultaneous counter for acquiring coincidence event data of a target region of a subject during a PET scan of the subject, and a processor, the processor comprising:

a first reconstruction unit for acquiring a PET image of a target region from the coincidence event data;

the time-of-flight processing unit is used for acquiring time difference between the timestamps of the coincidence events and classifying the coincidence event data into a plurality of boxes according to the time difference;

a correction unit for obtaining an initial scatter estimate for TOF-PET imaging based on the PET image and the coincidence event data classified into a plurality of bins; correcting the initial scattering estimation to obtain a scattering estimation of TOF-PET imaging; and scatter correcting the coincidence event data classified into bins based on scatter estimates of the TOF-PET imaging;

a second reconstruction unit for reconstructing the scatter-corrected coincidence event data classified into bins, acquiring a TOF-PET image of a target region.

Optionally, the correction unit is further configured to:

acquiring an anatomical image of a target region, wherein the anatomical image comprises classification information of a plurality of tissues;

registering the PET image to an anatomical image, and distributing corresponding attenuation values to voxels of the PET image according to the classification information of the anatomical image to obtain an attenuation map; and the number of the first and second groups,

performing attenuation correction on the TOF-PET data according to the attenuation map;

and/or, acquiring random coincidence event data of a target region of a subject and randomly correcting the TOF-PET data according to the random coincidence event data.

Compared with the prior art, the method has the following advantages: filtering the initial scattering estimation to obtain a smooth scattering vector and keep the shape of a scattering curve which can be obtained by the original TOF single scattering correction; fitting by using the fitting parameters reduces the quantitative difference between scattering correction acquired by TOF single scattering and subsequent fitting and scattering correction required by image consistency, and reduces the problem of accuracy reduction of reconstructed TOF-PET images caused by scattering estimation deviation; and the existing PET reconstruction is used as a reference, the physical correction of the TOF-PET data is further subjected to normalized correction, and the consistency of the TOF-PET image obtained by the TOF-PET reconstruction and the PET image obtained by the PET reconstruction is ensured.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. Like reference symbols in the various drawings indicate like elements.

FIG. 1 is a schematic diagram of a PET imaging system according to an embodiment of the present application;

FIG. 2 is a block diagram of a processor according to an embodiment of the present application;

FIG. 3 is a block diagram of a calibration unit according to an embodiment of the present application;

FIG. 4 is a flow chart of a PET imaging method according to an embodiment of the present application;

FIG. 5 is a schematic view of gamma photon scattering according to an embodiment of the present application;

FIG. 6A is a cross-sectional view of a PET image acquired according to an embodiment of the present application;

FIG. 6B is a cross-sectional view of a TOF-PET image obtained using a prior TOF-PET imaging method according to an embodiment of the present application;

FIG. 6C is a cross-sectional view of a TOF-PET image obtained using the method shown in FIG. 4 according to an embodiment of the present application;

FIG. 6D is a difference image acquired using the TOF-PET image of FIG. 6B and the PET image of FIG. 6A;

FIG. 6E is a difference image acquired using the TOF-PET image of FIG. 6C and the PET image of FIG. 6A;

FIG. 7A is a coronal view of a PET image acquired in accordance with an embodiment of the present application;

FIG. 7B is a coronal-plane view of a TOF-PET image obtained using a prior TOF-PET imaging method according to an embodiment of the present application;

FIG. 7C is a coronal-plane view of a TOF-PET image obtained using the method shown in FIG. 4 according to an embodiment of the present application;

FIG. 7D is a difference image acquired using the TOF-PET image of FIG. 7B and the PET image of FIG. 7A;

FIG. 7E is a difference image acquired using the TOF-PET image of FIG. 7C and the PET image of FIG. 7A;

FIG. 8A is a sagittal view of a PET image acquired according to an embodiment of the present application;

FIG. 8B is a sagittal view of a TOF-PET image obtained using a prior TOF-PET imaging method according to an embodiment of the present application;

FIG. 8C is a sagittal view of a TOF-PET image obtained using the method of FIG. 4 according to an embodiment of the present application;

FIG. 8D is a difference image acquired using the TOF-PET image of FIG. 8B and the PET image of FIG. 8A;

FIG. 8E is a difference image acquired using the TOF-PET image of FIG. 8C and the PET image of FIG. 8A.

Detailed Description

Some embodiments of the present application will be described below. It is noted that in the detailed description of these embodiments, in order to provide a concise description, all features of an actual implementation may not be described in detail.

It is obvious that the drawings in the following description are only examples or embodiments of the application, from which the application can also be applied to other similar scenarios without inventive effort for a person skilled in the art. Moreover, it should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions must be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which may vary from one implementation to another.

Unless otherwise defined, technical or scientific terms used in the claims and the specification should have the ordinary meaning as understood by those of ordinary skill in the art to which this application belongs. The use of "first," "second," and similar terms in the description and claims of this patent application do not denote any order, quantity, or importance, but rather the terms are used to distinguish one element from another. The terms "a," "an," "the," and the like, do not denote a limitation of quantity, and may denote the singular or plural.

The word "comprise" or "comprises", and the like, means that the element or item listed before "comprises" or "comprising" covers the element or item listed after "comprising" or "comprises" and its equivalent, and does not exclude other elements or items. "connected" or "coupled" and similar terms are not restricted to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

The aim of PET image reconstruction is mainly to reconstruct the spatial distribution of the contrast agent in the human body. At present, various methods can be adopted for PET image reconstruction, but activity mean values of PET images acquired by the methods at various positions should be consistent. Also, the activity of TOF-PET images obtained based on TOF-PET data reconstruction and conventional PET images should satisfy consistency, for example: a region of interest delineated in an organ at a non-pixel level should have the same activity mean in both the TOF-PET image and the PET image, and such a homogeneous activity image should not interfere with the diagnosis of the lesion by the physician.

The physical corrections of the TOF-PET reconstruction mainly include attenuation correction, randoms correction and scatter correction. Generally, attenuation correction is generally obtained by transforming a CT image registered with PET scanning through a bilinear model to obtain an attenuation image, projecting the attenuation image to a chordal image space and carrying out inverse logarithm acquisition; random events are typically acquired using a delayed time window; the scattering correction is difficult to correct, and especially when TOF-PET data are processed, the time-of-flight information and the sparser statistical data of each time-of-flight window further affect the scattering correction accuracy of each time window, so that the activity values of a TOF-PET image reconstructed based on the TOF-PET data and a conventional PET image have a consistency problem. The method takes the traditional PET (non-TOF PET) reconstruction as a reference or reference, and further carries out normalized correction on the physical correction of the TOF-PET data, so that the consistency of the TOF-PET image reconstructed by the TOF-PET data and the traditional PET image is realized.

The system of the present application may be used not only for non-invasive imaging such as diagnosis and research of diseases, but also in the industrial field, etc., and the image processing system thereof may include a positron emission computed tomography system (PET system), a positron emission computed tomography-computed tomography multi-modal system (PET-CT system), a positron emission computed tomography-magnetic resonance multi-modal hybrid system (PET-MR system), etc.

In one embodiment, a PET imaging system is illustrated. Referring to fig. 1, a diagram of a PET imaging system 100 according to an embodiment of the present application is illustrated. The PET imaging system 100 is centralized/controlled by a controller 110, having a couch, a gantry 120, a pre-processor 130, a coincidence counter 140, a processor 150, and an input/output device 160. The scan bed is used to support a subject and to move the subject (not shown in fig. 1) to a Field Of View (FOV) region.

The controller 110 may control the imaging process of the PET imaging system 100 including various data processing operations including attenuation correction, scatter correction, stochastic correction, and data reconstruction. It will be appreciated that the controller 110, the rack 120, the pre-processor 130, the coincidence counter 140, the processor 150, and the input/output device 160 may be interconnected. The connection may be a wireless network connection or a wired network connection. The wired network may include, among other things, utilizing one or more combinations of metallic cables, hybrid cables, optical fibers, one or more interfaces, and the like. The wireless network may include a network utilizing one or more combinations of bluetooth, Local Area Network (LAN), Wide Area Network (WAN), Near Field Communication (NFC), and the like. The controller 110 may be centralized, such as a data center; or may be distributed, such as a distributed system. The controller 110 may be local or remote.

In some embodiments, the controller 110 may include one or a combination of a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), an Application Specific Instruction Processor (ASIP), a Physical Processing Unit (PPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), a Programmable Logic Device (PLD), a microprocessor, a controller, a microcontroller, and the like.

The gantry 120 may be configured to be cylindrical, and a plurality of detector rings 121 are arranged along the central axis Z of the circumference, the detector rings 121 have a plurality of detectors/detectors arranged on the circumference of the central axis Z, and the subject may move with the scanning bed to the inside of the FOV of the scanning field enclosed by the plurality of detectors for imaging. In some embodiments, the PET imaging system 100 may provide only a single ring of radiation detectors; in other embodiments, the PET imaging system 100 includes two, three, four, five, or more rings that provide radiation detectors.

The detector has a scintillator (scintillator), a Photomultiplier Tube (PMT), and a light guide coupler. Among them, the scintillator may two-dimensionally arrange a plurality of crystals that can convert gamma rays incident from a subject into visible light.

In some embodiments, the scintillation crystal can be a single crystal or an array crystal, and the material of the scintillator can be NaI, BGO, LSO, LYSO, GSO, lutetium aluminate, lutetium pyrosilicate, or other lutetium compounds, or cerium-doped lanthanide, transition metal, or other inorganic crystals. The light guide coupler is generally made of plastic material with excellent light transmittance and is used for transmitting the visible light output by the scintillator to the photomultiplier; the photomultiplier tube receives visible light, multiplies the visible light, and converts the multiplied visible light into an electrical signal. The photomultiplier tube can be selected from Geiger Avalanche Photodiode (APD) or Si photomultiplier tube.

In one embodiment, the photomultiplier tube may select a series of APD cells, each cell being an independent geiger-type detector including a photoanode, a plurality of stages of dynodes, and a photocathode, wherein the photocathode receives scintillating light to produce photoelectrons; the multistage dynode may provide an electric field that accelerates photoelectrons, the electrons emitted from the photocathode are accelerated toward the dynode in the electric field and collide with the surface of the dynode to overflow a plurality of electrons, which phenomenon repeatedly occurs in the multistage dynode, and the number of electrons is avalanche-similarly multiplied; the photoanode can output millions of electrons to form an electron flow. It is noted that, in order to utilize the avalanche phenomenon for amplification, a voltage of 500-800 volts (V) can be applied between the dynode and the anode, and the gain range is 10⁵-10⁷Such that the detector has a sufficiently high temporal resolution to detect the time-of-flight difference between two "substantially simultaneous" gamma ray detections.

The preprocessor 130 is connected to the detector output end of the detector ring 121, and is configured to receive the electrical signal output by the detector and process the electrical signal. In one embodiment, a radioisotope-identified agent/tracer is injected into the subject prior to the PET scan; the detector detects the pair of annihilation gamma rays emitted from the inside of the subject, generates a pulse-like electric signal according to the light quantity of the detected pair of annihilation gamma rays, and supplies the pulse-like electric signal to the preprocessor 130; the preprocessor 130 generates Single Event Data (Single Event Data) from the electrical signal. The preprocessor 130 can detect the gamma rays that produce annihilations by detecting whether the intensity of the electrical signal exceeds a threshold. Alternatively, the preprocessor may employ Anger logic or other processing to identify the spatial coordinates, time stamp, and estimated energy of the detected gamma rays for a single event.

The coincidence counter 140 may be connected to an output of the preprocessor 130, and may receive the single-event data generated by the preprocessor 130 and perform coincidence counting processing on the single-event data related to a plurality of single events. Illustratively, the coincidence counter 140 repeatedly determines event data pertaining to two single events accommodated within a time range set in advance, for example, to about 6ns to 18ns, from the repeatedly supplied single event data. The paired single events are presumed to be due to pair annihilation gamma rays generated from the same pair of annihilation points, where the paired single events are broadly referred to as coincident events. The Line connecting the pair Of detectors that detect the pair Of annihilation gamma rays is called the Line Of Response (LOR), and the LOR data contains information about the number Of crystal bars in the detectors, the energy, the angle Of each LOR, and its distance from the center Of the FOV.

In one embodiment, each LOR may be numbered according to its angle and distance from the center of the FOV, and an array is constructed by the numbering, the elements in the array being the total number of coincidence events detected by the LOR, and this way of storing the coincidence events is a sine graph mode (Sinogram). Of course, coincidence event data can also be stored in a list mode (list mode), and the data in the list mode includes the number of crystal strips, photon energy, photon flight time information and the like. In this way, the coincidence counter 140 counts event data (hereinafter referred to as coincidence event data) regarding a coincidence event and a pair of events constituting an LOR for each LOR.

A processor 150 is connected to the output of the coincidence counter 140 for reconstructing image data from the coincidence event data, the image data representing a spatial distribution of the radioisotope concentration in the subject. In one embodiment, as shown in fig. 2, the processor 150 includes a time-of-flight processing unit 210 and a first reconstruction unit 220. The time-of-flight processing unit 210 can analyze the time difference (time-of-flight information) between the coincidence (response) event timestamps and classify the coincidence event data into bins (bins) based on the time difference to localize the positron-electron annihilation time along the LOR. Optionally, time-of-flight processing unit 210 locates the LOR within a distance interval corresponding to approximately the speed of light times the temporal resolution value of the radiation detector. The result of the accumulation of a large number of positron-annihilation events is a set of localized PET projection data.

The first reconstruction unit 220 may reconstruct a spatial distribution of the contrast agent in the subject based on the coincidence event data/PET projection data, and acquire a PET image corresponding to the PET projection data not containing time-of-flight information. In one embodiment, the first reconstruction unit 220 reconstructs coincidence event data through Filtered Back Projection (FBP), Maximum Expectation reconstruction (EM), Maximum A Posteriorit (MAP), and other algorithms to obtain a PET (activity) image/non-TOF PET image.

It is to be understood that the terms "PET image", "non-TOF PET image" and "non-TOF image" as used herein are intended to be synonymous and refer to the reconstruction of coincidence responsive event data by the conventional PET (not including time-of-flight techniques) reconstruction method to reproduce the tracer distribution in the human body.

In yet another embodiment, the processor 150 may include a correction unit 230, a second reconstruction unit 240, and/or a storage unit 250. The correction unit 230 may obtain scatter coincidence estimates during coincidence counting acquisition. In one embodiment, the PET image includes a plurality of voxels, and the correction unit may acquire an initial scatter estimate for TOF-PET imaging based on the PET image and coincidence event data classified into a plurality of bins; correcting the initial scattering estimation to obtain the scattering estimation of TOF-PET imaging; and scatter correcting the coincidence event data classified into the plurality of bins based on scatter estimates of the TOF-PET imaging.

It should be noted that, in the present application, the correction unit 230 obtains a scattering coincidence estimate in the coincidence counting acquisition process, which may be referred to as "first correction"; the initial scatter estimate is corrected to obtain a scatter estimate for TOF-PET imaging, which may be referred to as a "second correction". Optionally, the initial scattering estimate is corrected by a noise reduction method, a smoothing method, or other methods for filtering out signal noise.

In another embodiment, the correction unit 230 may acquire an anatomical image of the target region, the anatomical image containing classification information of a plurality of organs or tissues such as liver, lung, heart, fat, ribs, spine, and the like; registering the PET image to the anatomical image, assigning corresponding attenuation values/attenuation coefficients to voxels of the PET image according to classification information of the anatomical image, obtaining an attenuation map, and performing attenuation correction on the TOF-PET data based on the attenuation map. In yet another embodiment, the correction unit 230 may acquire random coincidence event data of a target region of a subject and randomly correct the TOF-PET data according to the random coincidence event data. Alternatively, the anatomical image may be a CT image, and the voxels of the PET image may be assigned respective attenuation values/attenuation coefficients by the CT values of the CT image.

The second reconstruction unit 240 may reconstruct the scatter-corrected coincidence event data classified into the plurality of bins, acquiring TOF-PET/TOF images of the target region. In one embodiment, second reconstruction unit 240 may obtain scatter-corrected coincidence event data classified into bins from correction unit 230; and reconstructing the data by adopting an analytical method. In another embodiment, the second reconstruction unit 240 may employ a MLEM reconstruction algorithm to reconstruct the scatter-corrected coincidence event data classified into a plurality of bins. In yet another embodiment, the second reconstruction unit 240 may further obtain a TOF-PET image with high image resolution by using a method based on a combination of a system response Function (PSF) and TOF-PET reconstruction.

It is understood that "TOF-PET reconstruction" and "TOF reconstruction" as used herein can be used in the same sense, and all refer to assigning coincidence event data to LORs corresponding to temporal resolution according to different weights based on time-of-flight information, and performing image reconstruction on the coincidence event data to obtain distribution images of tracers. Similarly, "TOF-PET image" and "TOF image" may be used in the same sense, both referring to the distribution image of the tracer within the organism.

Storage unit 250 may be provided as a device for data storage, such as a floppy disk drive, an optical disk, a ROM (read only memory), a RAM (random access memory), an EPROM (erasable programmable read only memory), an EEPROM (electrically erasable programmable read only memory), a CD-ROM drive (compact disk-read only memory), a DVD drive, a magneto-optical disk, or any other digital device for reading instructions and/or data from a computer-readable medium. Optionally, the instructions executed by the processor and/or the processed data are stored in a medium such as a hard disk or a cloud storage.

An input/output device 160 may be coupled to an output of the processor 150 for displaying the processor-generated PET image and/or TOF-PET image, with the PET image having consistency with the TOF-PET image. Illustratively, the input/output devices 160 may include a mouse, a keyboard, a display, and a human interaction device. In one embodiment, the display displays the subject's height, weight, age, imaging location, and operating state of the PET imaging system 100, among other things. The type of display may be one or a combination of several of a Cathode Ray Tube (CRT) display, a Liquid Crystal Display (LCD), an Organic Light Emitting Display (OLED), a plasma display, and the like.

Fig. 3 is a schematic structural diagram of a calibration unit 230 according to an embodiment of the present application. The correction unit 230 may include an initial scatter estimate acquisition subunit 310, a high frequency component filtering subunit 320, a fitting parameter acquisition subunit 330, and a TOF-PET imaging scatter estimate acquisition subunit 340.

The initial scatter estimate acquisition subunit 310 may acquire a PET image from the first reconstruction unit 220, TOF-PET data from the coincidence counter 140, and an initial scatter estimate for TOF-PET imaging based on the PET image and the TOF-PET data. The high frequency component filtering subunit 320 may acquire an initial scatter estimate for TOF-PET imaging from the initial scatter estimate acquiring subunit 310 and perform a correction process on the initial scatter estimate. The fitting parameter acquiring subunit 330 may acquire a reference of TOF-PET scattering using a time-of-flight based single correction method; and fitting the initial scattering estimation according to the standard of TOF-PET scattering to obtain fitting parameters. The TOF-PET imaging scatter estimate acquisition subunit 340 may acquire fitting parameters from the fitting parameter acquisition subunit 330 and acquire scatter estimates for TOF-PET imaging based on the fitting parameters and a baseline for TOF-PET scatter. The fitting parameter obtaining subunit 330 may further perform the following operations: (quadratic) fitting the initial scatter estimate with TOF scatter estimates mutually consistent with the PET image to obtain fitting parameters. It will be appreciated that determining TOF scatter estimates that are mutually consistent with the PET image may be referred to as a "first-fit" process, while fitting the initial scatter estimates using TOF scatter estimates that are mutually consistent with the PET image may be referred to as a "second-fit" process.

It should be noted that, in the present application, the initial scattering estimation obtaining subunit 310 obtains the scattering coincidence estimation in the coincidence counting acquisition process, which may be referred to as "first correction" of the scattering event/data; the high frequency component filtering subunit 320 performs a correction process on the initial scatter estimate to obtain a scatter estimate for TOF-PET imaging, which may be referred to as a "second correction" of the scatter events/data. Optionally, the initial scattering estimate is corrected by a noise reduction method, a smoothing method, or other methods for filtering out signal noise.

As shown in fig. 4, a PET imaging method according to an embodiment of the present application is used for TOF-PET imaging, and includes the following steps:

during a PET scan of a subject, TOF-PET data of a target region of the subject is acquired, the TOF-PET data including coincidence event data and TOF information, wherein the coincidence event data and/or the TOF information can be stored in a sinogram mode. Optionally, the coincidence event data may include true coincidence data without including scatter coincidence event data and/or random coincidence event data.

In some embodiments of the present application, during a scan using the PET imaging system 100 shown in fig. 1, a positron radionuclide previously injected into the subject emits positrons as it decays, while protons in the nuclei decay into neutrons and release positrons and neutrals, the positrons being equal in mass and quantity to electrons, and the electrons being equal in quantity and opposite in sign. The positron travels 1-3mm within the body tissue before annihilation, producing 511keV gamma photons that are 180 ° apart. Two oppositely directed gamma photon rays hit the detectors in the detector ring 121, which are in the same line, almost simultaneously and generate an electrical signal/coincidence trigger signal. The electrical signal is converted to single-event data by the preprocessor 130 while the counter 140 receives the single-event data sent by the preprocessor 130 and determines whether two events within a predetermined time window are the same annihilation event. If the same annihilation event is present, the coincidence event data corresponding to the line of coincidence response, which may also be referred to as PET projection data, is concurrently counted by the counter 140 to initiate a line of response (LOR) processor to determine the line of response along which the annihilation event occurred.

FIG. 5 is a schematic diagram of coordinates of a PET scanning cavity according to an embodiment of the present invention (the center of the cavity enclosed by the frame 12 or the detector ring 121 is taken as the origin of coordinates, the radius of the circumference of the longitudinal section of the cavity is r, a plurality of annular detectors are distributed around the cavity, and two photons emitted in a coincidence event can be detected, for example, in an X-Y plane (the cross section of PET is parallel to the plane of the detection ring), a detector A (X is parallel to the plane of the detection ring), and a detector A (X is_a,y_a) And a detector B (x)_b,y_b) For two crystal detectors hit by a true coincidence event, the line connecting the two detectors passes through the scattering point, and the line is a response line.

In this embodiment, a connection line between the detector a and the scattering point S of the photon H is a response line AS, and a connection line between the detector B and the scattering point S is a response line BS. In the absence of dielectric scattering, the flight path of photon H may move along the line of response AS and CS with a 180 degree phase shift from line of response AS, respectively. However, due to the presence of medium scattering, the obtained response line is not completely matched with the actual flight path of the photon, and the flight path of the photon H can respectively follow the response lineAS and response line BS, the phase shift between the two response lines is less than 180 degrees; in this case the line connecting detector A and detector B is considered to correspond to line of response LOR_ABAnd the position of the scattering point determined according to the response line is not completely consistent with the actual position, so that the finally reconstructed PET image has certain deviation from the true distribution of the tracer.

In another embodiment, time-of-flight TOF information of two gamma ray detection events/coincidence events arriving at detector a and detector B substantially simultaneously may be detected and the time-of-flight TOF information/data may be represented by a probability density function.

Step 402 the first reconstruction unit 220 acquires a PET image of the target region from the coincidence event data, the PET image comprising a plurality of voxels. The relationship of coincidence event data and PET (non-TOF PET) image/activity image can be represented by the following model:

wherein, y_iThe value of the ith element of the acquired coincidence event data (chordal pattern), which may also be referred to as PET projection data; a is_iA value representing the ith element of the attenuation correction chord chart; p_ijRepresenting the value of the jth image element and ith chord chart element corresponding to a system matrix; x is the number of_jA value representing the jth element of the PET image; r is_iAnd s_iValues representing the random correction and scatter correction ith element.

The PET image may be obtained by performing multiple iterations of the acquired PET data using an ordered-subset maximum expected value (OS-EM) reconstruction algorithm, which may be represented, for example, by the following equation for any one of the iterations of the reconstruction process:

wherein,a value representing the jth element in the reconstructed PET image updated by the mth subset of the nth iteration;a value representing the jth element in the reconstructed PET image before the mth subset is updated by the nth iteration;a value representing the kth element in the reconstructed PET image before the mth subset update is performed on the nth iteration; y is_iA value representing the ith element of the acquired PET projection data; a is_iA value representing the ith element of the attenuation correction chord chart; p_ikRepresenting the values, P, of a system matrix corresponding to the kth image element, the ith chord element_ijRepresenting the value of the jth image element and ith chord chart element corresponding to a system matrix; r is_iAnd s_iValues representing the random correction and scatter correction ith elements, respectively; y is_iRepresenting the value of the ith element of the acquired PET projection data. A PET image/non-TOF PET image reflecting the radionuclide decay profile of the subject's target region, which contains a plurality of voxels and does not utilize time-of-flight information, is obtained through an iterative process similar to that described above.

In step 403, the initial scatter estimate acquisition subunit 310 acquires an initial scatter estimate for TOF-PET imaging based on the PET image and the TOF-PET data. In one embodiment, the initial scatter estimate for TOF-PET imaging may be obtained by the following equation:

wherein t represents the number of bins of time of flight;respectively representing scatter correctionThe value of the t-th time-of-flight bin, the i-th element;representing values corresponding to the t-th time-of-flight bin and the i-th element in TOF-PET data; a is_iA value representing the ith element of the attenuation correction chord chart;representing values of a system matrix corresponding to a jth image element and an ith chord map element in a tth time-of-flight bin;representing PET data corresponding to jth voxel in the PET image or reconstructed PET data;the values corresponding to the t-th time-of-flight bin and the i-th element in the random correction are shown.

And step 404, correcting the initial scattering estimation to obtain a scattering estimation of TOF-PET imaging. Typically, the scatter correction sinogram should spatially contain low frequency signals, and the calculation according to equation (3) directly introduces high frequency components in the data into the scatter correction, resulting in insignificant advantages of TOF reconstruction over conventional PET reconstruction. Considering that the scatter correction should be smooth, the present application in one embodiment performs a high frequency filtering process on the initial scatter estimate by the high frequency component filtering subunit 320 to remove the high frequency components contained in the initial scatter estimate. Illustratively, the following formula may be employed:

wherein L denotes a filter. Optionally, the type of the filter is selected from a physical model like a gaussian shape such as a butterworth low-pass filter, an adaptive low-pass filter, etc., to obtain the TOF imaging scatter estimate without high frequency components. The high frequency components of the preliminary scatter estimate are removed by applying a filter to obtain a signal containing only low frequency components, which can be used to represent the scatter event estimate.

In another embodiment, a smooth scattering chord chart obtained by a Single Scattering Simulation (SSS) method based on time of flight is used as a reference for TOF-PET scattering, and initial TOF-PET scattering (formula (3)) obtained by the TOF Single calibration method is used for fitting the reference TOF scattering to obtain a fitted smooth scattering chord chart.

Optionally, the scatter estimate for TOF-PET imaging is obtained by: the fitting parameter acquiring subunit 330 acquires a reference of TOF-PET scattering by using a single correction method based on time of flight; fitting the initial scattering estimation according to the TOF-PET scattering reference to obtain fitting parameters; the TOF-PET imaging scatter estimate acquisition subunit 340 acquires scatter estimates for TOF-PET imaging based on the fitting parameters and a reference for TOF scatter. Illustratively, the following formula may be employed:

wherein t represents the bin number of the time of flight; a denotes one detector detecting a coincident event photon, B denotes another detector detecting a coincident event photon, (a, B) denotes a detector pair of coincident events; s represents a scattering point, S^ABIs a non-dimensional quantity representing the count rate of a single scattered photon pair from a scattering point detected by the (A, B) detector pair; s^AB(ray, angle, slice, t) represents a four-dimensional sinogram which can be used as a form standard of scattering, and ray, angle, slice, t respectively represent one-dimensional projection, angle, bedding and flight time box number and correspond to the ith element of the tth flight time box;andrepresenting a parameter related to scatter correction; s_finalRepresenting TOF-PET imaging scatter estimates; tail (t, slice) represents a time-of-flight box numbered t, a trailing region in the two-dimensional chord graph under the slice level; min represents the operation of taking the first function. The method effectively reduces the quantitative difference of scattering correction required by the consistency of scattering correction acquired by the TOFSSS and subsequent fitting and the image on the premise of ensuring the smoothness of the acquired scattering estimation vector, and reduces the problem of low precision of the TOF-PET image caused by scattering estimation deviation.

In one embodiment, the TOFFSSS algorithm for single scatter correction of TOF-PET data can be expressed by the following equation:

let AS denote the line of response from scatter point S to detector a and BS denote the line of response from scatter point S to detector B; order toRepresents the component of the detector a corresponding to the time-of-flight bin numbered t; order toRepresents the component of the detector B corresponding to the time-of-flight bin numbered t; sigma_ASRepresenting the geometric cross-section of the A-probe corresponding to the line of response AS，σ_BSRepresenting the geometric cross section of the B detector corresponding to the response line BS; r_ASRepresenting the distance, R, from the scattering point S to the detector A_BSRepresents the distance from the scattering point S to the detector B; let μ denote the attenuation coefficient corresponding to the 511keV energy level, and μ' denote the attenuation coefficient of the scattered photon; sigma_CRepresents a cross-section of a compton contact; ω represents the scattering solid angle; order to_ASRepresents the detection efficiency at 511keV energy level of detector A corresponding to the response line AS'_ASThe detection efficiency of the detector A corresponding to the response line AS is shown when the energy corresponding to the scattered photons is obtained; order to_BSRepresenting the detection efficiency at 511keV energy level of detector B corresponding to the line of response BS'_BSRepresenting the detection efficiency of the detector B corresponding to the response line BS when the energy corresponding to the scattered photons is obtained; ρ represents the radioactivity distribution activity.

In yet another embodiment, the correction processing of the initial scatter estimate to obtain a scatter estimate for TOF-PET imaging comprises: performing (quadratic) fitting on the initial scattering estimation by using TOF scattering estimation mutually consistent with the PET image to obtain fitting parameters; a scatter estimate for TOF-PET imaging is obtained based on the fitting parameters and the initial scatter estimate. Illustratively, the fitting parameter obtaining subunit 330 obtains TOF scatter estimates that are mutually consistent with the PET image based on a monte carlo model; fitting the initial scattering estimation according to TOF scattering estimation mutually consistent with the PET image to obtain fitting parameters; the TOF-PET imaging scatter estimate acquisition subunit 340 acquires scatter estimates for the TOF-PET imaging based on the fitting parameters and the TOF-PET scattered reference.

It should be noted that, in the present application, the angle, the layer, and each dimension under the time flight box of the scattering chord chart are projected, and the corresponding scattering vector is still smooth, and the shape of the scattering curve obtained by the original toffsss is maintained; and fitting the form benchmark of the scattering estimation by adopting the fitting parameters, and further reducing the quantitative difference of the scattering correction required by consistency/mutual consistency of the scattering correction acquired by the TOFSSS and subsequent fitting and the image.

It will be appreciated that in the above process, obtaining an estimate of the scatter coincidences during the coincidence counting acquisition process may be referred to as a "first correction" of the scatter events/data; the initial scatter estimate is corrected to obtain a scatter estimate for TOF-PET imaging, which may be referred to as a "second correction" of the scatter events/data. Of course, the initial scattering estimate is corrected by a noise reduction method, a smoothing method, or other methods for filtering out signal noise.

Step 405. performing scatter correction on the TOF-PET data based on the scatter estimate of the TOF-PET imaging; and reconstructing the scatter-corrected TOF-PET data to obtain a TOF-PET image of the target region. In one embodiment, the second reconstruction unit 240 acquires TOF-PET images using an ordered subset maximum expected value (OS-EM) reconstruction algorithm, which illustratively may use the following formula:

wherein, let a_iThe value representing the ith element in the attenuation sinogram may also be referred to as the attenuation value; order toRepresenting the values/elements in the system matrix corresponding to the jth image element, ith chord element in the time-of-flight box numbered t,representing the values/elements in the system matrix corresponding to the kth image element, the ith chord element in the time-of-flight box numbered t; r is_iA value representing a random correction of the ith element; s_iA value representing the ith element in a scatter estimate for TOF imaging; order toRepresenting the value of the jth element/pixel point in the reconstructed image updated by the mth subset of the nth iteration;a value representing the ith element in the time-of-flight bin numbered t in the TOF-PET data; order toAnd representing the value of the jth element/pixel point in the reconstructed image before the mth subset is updated through the nth iteration. In the embodiment, the TOF-PET scattering estimation is close to the true value, so that the TOF-PET image accuracy obtained by reconstruction can be effectively improved.

Optionally, attenuation correction and/or random correction may also be performed on the TOF-PET data during reconstruction of the TOF-PET data into a TOF-PET image. In one embodiment, attenuation correction of TOF-PET data may include the steps of: acquiring an anatomical image corresponding to a target region, wherein the anatomical image can be a CT image, an MR image and other anatomical images, and the anatomical image can contain classification information of a plurality of tissues such as lung, fat, ribs, spine, heart and the like; registering the PET image to an anatomical image, and distributing corresponding attenuation values to voxels of the PET image according to the classification information of the plurality of tissues to obtain an attenuation map; and performing attenuation correction during reconstruction of the scatter-corrected TOF-PET data into a TOF-PET image of the region of interest based on the attenuation map.

Fig. 6A, 6B and 6C are cross sectional views of a PET image acquired according to an embodiment of the present application, a TOF-PET image obtained using a conventional PET imaging method, and a TOF-PET image obtained using the method shown in fig. 4, respectively. Whether TOF-PET images and nonTOF-PET imaging reconstruction images have larger difference or not can be determined by respectively comparing TOF-PET images obtained by the existing TOF-PET imaging method and TOF-PET images obtained by the method with PET images. Fig. 6D is a difference image obtained by using the TOF-PET image of fig. 6B and the PET image of fig. 6A, wherein the black pixel region indicates that the TOF-PET reconstructed pixel value is higher than the non-TOF-PET reconstructed pixel value; the white pixel area indicates that the pixel values of the TOF-PET reconstruction are lower than those of the nonTOF-PET reconstruction. Further, the image difference exhibits a tendency to vary with spatial position variation. The pixels in the middle area of fig. 6D are black areas, the gray values of the portions closer to the skin line or the organ boundary are smaller, and the boundary areas of the two types of pixels have obvious jump, which indicates that there is a large amount of inconsistency between the pixels of the image obtained by the TOF-PET method and the non-TOF-PET reconstruction method. As shown in fig. 6E, compared with the TOF-PET image obtained by the PET imaging method of the present application, the TOF-PET image has more uniform pixel values in the whole area, and the two types of pixel points have less obvious jump in the cross-over area, so that the uniformity of the pixel values of the pixel points in the whole TOF-PET image is improved.

Fig. 7A, 7B and 7C are respectively a coronal view of a PET image acquired according to an embodiment of the present application, a coronal view of a TOF-PET image acquired by using a conventional PET imaging method, and a coronal view of a TOF-PET image acquired by using the method shown in fig. 4. The difference image shown in fig. 7D can be obtained by subtracting the pixel values of the voxels in fig. 7B and 7A, wherein the upper and lower parts of the difference image have more black pixel points, the middle part has more white pixel points, and the pixel values of the two parts are more obviously inconsistent. Subtracting the pixel values of each voxel in fig. 7C and 7A can obtain a difference image as shown in fig. 7E, which has more uniform pixel values of each voxel than fig. 7D and has better consistency with the corresponding portion in fig. 6E.

Fig. 8A, 8B and 8C are respectively a sagittal view of a PET image acquired according to an embodiment of the present application, a sagittal view of a TOF-PET image acquired by using a conventional PET imaging method, and a sagittal view of a TOF-PET image acquired by using the method shown in fig. 4. The difference image shown in fig. 8D can be obtained by subtracting the pixel values of the voxels in fig. 8B and 8A, where the upper and lower portions of the difference image corresponding to fig. 7D have more black pixel points, the middle portion has more white pixel points, and the pixel values of the two portions are not consistent obviously. Subtracting the pixel values of each voxel in fig. 8C and 8A can obtain a difference image as shown in fig. 8E, where the pixel values of each voxel in the image are more uniform than those in fig. 8D, and have better consistency with the corresponding parts in fig. 6E and 7E, and the pixel inconsistency between the TOF-PET image and the conventional PET image is improved.

Having thus described the basic concept, it will be apparent to those skilled in the art that the foregoing disclosure is by way of example only, and is not intended to limit the present application. Various modifications, improvements and adaptations to the present application may occur to those skilled in the art, although not explicitly described herein. Such modifications, improvements and adaptations are proposed in the present application and thus fall within the spirit and scope of the exemplary embodiments of the present application.

Also, this application uses specific language to describe embodiments of the application. Reference throughout this specification to "one embodiment," "an embodiment," and/or "some embodiments" means that a particular feature, structure, or characteristic described in connection with at least one embodiment of the present application is included in at least one embodiment of the present application. Therefore, it is emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, some features, structures, or characteristics of one or more embodiments of the present application may be combined as appropriate.

Computer program code required for the operation of various portions of the present application may be written in any one or more programming languages, including an object oriented programming language such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C + +, C #, VB.NET, Python, and the like, a conventional programming language such as C, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, a dynamic programming language such as Python, Ruby, and Groovy, or other programming languages, and the like. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any network format, such as a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet), or in a cloud computing environment, or as a service, such as a software as a service (SaaS).

Additionally, the order in which elements and sequences of the processes described herein are processed, the use of alphanumeric characters, or the use of other designations, is not intended to limit the order of the processes and methods described herein, unless explicitly claimed. While various presently contemplated embodiments of the invention have been discussed in the foregoing disclosure by way of example, it is to be understood that such detail is solely for that purpose and that the appended claims are not limited to the disclosed embodiments, but, on the contrary, are intended to cover all modifications and equivalent arrangements that are within the spirit and scope of the embodiments herein. For example, although the system components described above may be implemented by hardware devices, they may also be implemented by software-only solutions, such as installing the described system on an existing server or mobile device.

Similarly, it should be noted that in the preceding description of embodiments of the application, various features are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure aiding in the understanding of one or more of the embodiments. This method of disclosure, however, is not intended to require more features than are expressly recited in the claims. Indeed, the embodiments may be characterized as having less than all of the features of a single embodiment disclosed above.

Claims

1. A PET scatter correction method, comprising:

acquiring a PET image of a target area according to the coincidence event data;

2. The PET scatter correction method of claim 1, wherein the correction processing of the initial scatter estimate comprises: and performing high-frequency filtering processing on the initial scattering estimation to remove high-frequency components contained in the initial scattering estimation.

3. The PET scatter correction method of claim 1, wherein the correction processing of the initial scatter estimate to obtain a scatter estimate for TOF-PET imaging comprises:

4. The PET scatter correction method of claim 1, wherein the correction processing of the initial scatter estimate to obtain a scatter estimate for TOF-PET imaging comprises:

obtaining a scatter estimate for TOF-PET imaging based on the fitting parameters and the initial scatter estimate.

5. A PET imaging method, comprising:

acquiring a PET image of a target area according to the coincidence event data;

6. The PET imaging method of claim 5 wherein the correction processing of the initial scatter estimate to obtain a scatter estimate for TOF-PET imaging comprises:

7. The PET imaging method according to claim 5 wherein the coincidence event data, the TOF information and/or the TOF-PET data are stored in a sinogram mode.

8. The PET imaging method according to claim 7, further comprising: