CN109124672A - Bearing calibration, device, equipment and the storage medium of random coincidence event - Google Patents
Bearing calibration, device, equipment and the storage medium of random coincidence event Download PDFInfo
- Publication number
- CN109124672A CN109124672A CN201811224562.9A CN201811224562A CN109124672A CN 109124672 A CN109124672 A CN 109124672A CN 201811224562 A CN201811224562 A CN 201811224562A CN 109124672 A CN109124672 A CN 109124672A
- Authority
- CN
- China
- Prior art keywords
- event
- delay
- packet loss
- string
- pet detector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000013135 deep learning Methods 0.000 claims abstract description 46
- 238000000034 method Methods 0.000 claims abstract description 36
- 238000004590 computer program Methods 0.000 claims description 30
- 238000012545 processing Methods 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 6
- 238000002059 diagnostic imaging Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 abstract description 10
- 238000012937 correction Methods 0.000 abstract description 10
- 230000000694 effects Effects 0.000 abstract description 8
- 238000012549 training Methods 0.000 description 14
- 230000008439 repair process Effects 0.000 description 7
- 230000005540 biological transmission Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000001514 detection method Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000011084 recovery Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000003111 delayed effect Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 239000000523 sample Substances 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002285 radioactive effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
- A61B6/03—Computed tomography [CT]
- A61B6/037—Emission tomography
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5205—Devices using data or image processing specially adapted for radiation diagnosis involving processing of raw data to produce diagnostic data
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5258—Devices using data or image processing specially adapted for radiation diagnosis involving detection or reduction of artifacts or noise
Landscapes
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Pathology (AREA)
- Heart & Thoracic Surgery (AREA)
- High Energy & Nuclear Physics (AREA)
- Physics & Mathematics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Optics & Photonics (AREA)
- Veterinary Medicine (AREA)
- Radiology & Medical Imaging (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Nuclear Medicine (AREA)
Abstract
This application involves bearing calibration, device, equipment and the storage mediums of a kind of random coincidence event.The described method includes: pet detector acquisition meets the delay event met in event and delay time window in time window, packet loss operation is carried out to delay event, obtain section retards event and packet loss, event will be met, section retards event and packet loss are sent to image reconstruction processor, section retards event is rearranged to string figure by image reconstruction processor, according to the string figure and packet loss of section retards event, the string figure of all delay events of pet detector acquisition is obtained by deep learning network, according to the string figure of all delay events, the random coincidence event met in event is corrected.Pet detector can reduce for the collected delay event that meets when event and delay event export to the occupancy of communication bandwidth using this method, it avoids the problem that delay event excessively occupies communication bandwidth and causes to meet event loss, improves the effect of random coincidence event correction.
Description
Technical field
This application involves PET technical field of imaging, more particularly to a kind of bearing calibration of random coincidence event, device,
Equipment and storage medium.
Background technique
When PET/CT or PET/MR is imaged, what detector acquired in meeting window, which meets event, to be divided into three classes:
True coincidence event (True events), scattering meet event (Scatters) and random coincidence event (Random events).
The increase of random coincidence event will increase the noise of PET image, influence the contrast of PET image, in the school of random coincidence event
Center, most common method is delay window method.Delay window method meet time window (such as 4ns) close after by it is sufficiently long when
Between (such as 100ns) after, it opens a delay time window and carries out data acquisition, the data acquired in delay time window, which are referred to as, prolongs
Slow event (Delay events), so that delay event is all to belong to random coincidence event.Since random coincidence event is in the time
There is random distribution nature in dimension, it can be using the delay event in delay time window to the random coincidence met in time window
Event is estimated.
In routine clinical PET scan, true coincidence event, scattering meet event and random coincidence event these three types event
Accounting respectively may be about 30%, 30% and 40%.As radioactive source activity increases, accounting of the random coincidence event in all events
It will steeply rise, this meaning will have a large amount of communication bandwidth to be used for the biography of delay event in delay window method in high activity
It is defeated, so that communication bandwidth is consumed totally quickly.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of bearing calibration of random coincidence event, device, set
Standby and storage medium, to solve meeting the random coincidence event in time window by the delay event estimation in delay time window
When, delay event is higher to the consumption of communication bandwidth, causes random coincidence event correction ineffective, and then influence reconstruction image
The problem of quality.
A kind of bearing calibration of random coincidence event, which comprises
Pet detector acquisition meets the delay event met in event and delay time window in time window；
The pet detector carries out packet loss operation to the collected delay event, obtains in the delay event
Section retards event and packet loss send out the packet loss for meeting event, the section retards event and the delay event
It send to image reconstruction processor；
The section retards event is rearranged to string figure by described image reconstruction processor, and according to the section retards event
Corresponding string figure and the packet loss obtain the institute of the pet detector acquisition by preparatory trained deep learning network
There is the corresponding string figure of delay event；
The corresponding string figure of all delay events that described image reconstruction processor is acquired according to the pet detector, to institute
The random coincidence event met in event is stated to be corrected.
In some embodiments, the step of pet detector carries out packet loss operation to the collected delay event,
Include:
Packet loss operation is carried out to the delay event according to specified time interval.
In some embodiments, the step of pet detector carries out packet loss operation to the collected delay event,
Include:
Packet loss operation is carried out to the collected delay event according to default packet loss.
In some embodiments, according to the corresponding string figure of the section retards event and the packet loss, by instructing in advance
The deep learning network perfected obtains the step of all delay events corresponding string figure of pet detector acquisition, further includes:
Data normalization processing is carried out to the corresponding string figure of the section retards event, it will treated section retards
The corresponding string figure of event and the packet loss are input in the deep learning network.
In some embodiments, the step of pet detector carries out packet loss operation to the collected delay event,
Further include:
Obtain the current counting rate of the pet detector, according to the counting rate to the packet loss of the delay event into
Mobile state adjustment；
Packet loss operation is carried out to the delay event according to the packet loss of dynamic adjustment.
In some embodiments, according to the corresponding string figure of the section retards event and the packet loss, by instructing in advance
The deep learning network perfected obtains the step of all delay events corresponding string figure of pet detector acquisition, further includes:
The average packet loss ratio of the delay event is calculated according to the packet loss that the dynamic adjusts；
The corresponding string figure of the section retards event and the average packet loss ratio are input in the deep learning network,
Generate the corresponding string figure of all delay events of the pet detector acquisition.
In some embodiments, the method also includes:
Whether the current count rate for judging the pet detector is more than preset counting rate threshold value；
When the current count rate of the pet detector is less than the counting rate threshold value, meet event and institute for described
It states delay event and is sent to image reconstruction processor；
All delay events that described image reconstruction processor acquires the pet detector are rearranged to string figure, according to institute
The corresponding string figure of delay event is stated, the random coincidence event met in event is corrected.
A kind of medical imaging apparatus, described device include:
Pet detector, for acquiring the delay event met in event and delay time window met in time window, and it is right
The delay event carries out packet loss operation, obtains the section retards event and packet loss in the delay event, meets described
The packet loss of event, the section retards event and the delay event is sent to image reconstruction processor；And
Described image reconstruction processor for the section retards event to be rearranged to string figure, and prolongs according to the part
It is corresponding to obtain the delay event by preparatory trained deep learning network for the slow corresponding string figure of event and the packet loss
String figure the random coincidence event met in event is corrected according to the corresponding string figure of the delay event.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the processing
Device performs the steps of when executing the computer program
Pet detector acquisition meets the delay event met in event and delay time window in time window；
The pet detector carries out packet loss operation to the collected delay event, obtains in the delay event
Section retards event and packet loss send out the packet loss for meeting event, the section retards event and the delay event
It send to image reconstruction processor；
The section retards event is rearranged to string figure by described image reconstruction processor, and according to the section retards event
Corresponding string figure and the packet loss obtain the institute of the pet detector acquisition by preparatory trained deep learning network
There is the corresponding string figure of delay event；
The corresponding string figure of all delay events that described image reconstruction processor is acquired according to the pet detector, to institute
The random coincidence event met in event is stated to be corrected.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
It is performed the steps of when row
Pet detector acquisition meets the delay event met in event and delay time window in time window；
The pet detector carries out packet loss operation to the collected delay event, obtains in the delay event
Section retards event and packet loss send out the packet loss for meeting event, the section retards event and the delay event
It send to image reconstruction processor；
The section retards event is rearranged to string figure by described image reconstruction processor, and according to the section retards event
Corresponding string figure and the packet loss obtain the institute of the pet detector acquisition by preparatory trained deep learning network
There is the corresponding string figure of delay event；
The corresponding string figure of all delay events that described image reconstruction processor is acquired according to the pet detector, to institute
The random coincidence event met in event is stated to be corrected.
Bearing calibration, device, equipment and the storage medium of above-mentioned random coincidence event, pet detector is by collected symbol
When conjunction event and delay event are transferred to image reconstruction processor, by carrying out packet loss operation to delay event, delay thing is reduced
Part avoids the problem that delay event excessively occupies communication bandwidth and causes to meet event loss to the occupancy of communication bandwidth,
Image reconstruction processor receive meet the section retards event obtained after event, packet loss and postpone event packet loss after,
Section retards event is repaired by deep learning network, all delay events of pet detector acquisition is obtained, passes through this
A little delay events are corrected the random coincidence event met in event, to improve the effect of random coincidence event correction, into
And improve PET image reconstruction effect.
Detailed description of the invention
Fig. 1 is the flow diagram of the bearing calibration of random coincidence event in some embodiments；
Fig. 2 is the structural block diagram of some embodiment traditional Chinese medicine imaging devices；
Fig. 3 is the structural block diagram of the means for correcting of random coincidence event in some embodiments；And
Fig. 4 is the internal structure chart of computer equipment in some embodiments.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
In some embodiments, as shown in Figure 1, providing a kind of bearing calibration of random coincidence event, including following step
It is rapid:
Step 102, pet detector acquisition meets the delay event met in event and delay time window in time window.
Wherein, in PET scan system, radionuclide occurs mutually to turn between proton and neutron in decay process
Change, launches the positive electron with certain kinetic energy, positive electron (i.e. positive negative electricity in conjunction with negative electron around after flight certain distance
Son is buried in oblivion), launch two photons that energy is identical, contrary, if this to photon it is same meet it is a pair of in time window
PET probe unit is detected simultaneously by, and obtains a pulse signal just to get to the true coincidence event met in event.Due to one
To PET probe unit, the photon that detects may not be the pulse signal that obtains at this time from the same annihilation event simultaneously
To meet the random coincidence event in event.Random coincidence event is one of main noise source during PET scan, therefore
It needs to be corrected the random coincidence event met in event.
Specifically, pet detector acquisition meets the event that meets in time window, and when generating this according to delay window method and meeting
Between the corresponding delay time window of window, acquire delay time window in delay event.
Step 104, pet detector carries out packet loss operation to collected delay event, obtains the part in delay event
The packet loss for meeting event, section retards event and delay event is sent to image reconstruction process by delay event and packet loss
Device.
Specifically, when pet detector will meet event and delay event and transmit to image reconstruction processor, to delay thing
Part carries out packet loss operation, the delay event after obtaining packet loss.In order to be distinguished with the delay event before packet loss, herein by packet loss
Delay event afterwards is known as section retards event.Losing for event, section retards event and delay event will be met by pet detector
Packet rate is sent to image reconstruction processor.
In some embodiments, pet detector is by different transmission channels to meeting event and delay event passes
It is defeated, individually to carry out the integrality that packet loss is operated without influencing to meet event to delay event.
Step 106, section retards event is rearranged to string figure by image reconstruction processor, and corresponding according to section retards event
String figure and packet loss, all delay events pair of pet detector acquisition are obtained by preparatory trained deep learning network
The string figure answered.
Specifically, meet event, section retards event and the packet loss for postponing event when image reconstruction processor receives
When, the section retards event is reset, the corresponding string figure of the section retards event is obtained.Section retards event is corresponding
String figure and packet loss input trained deep learning network, to be carried out by deep learning network to the section retards event
Repairing, obtains the string figure of deep learning network output, and all delay events that the string figure of the output is acquired by pet detector are right
The string figure answered, to complete the recovery to event is postponed after packet loss.
Wherein, deep learning network can be convolutional neural networks, residual error network etc..
Wherein, string figure can be sinogram (sinogram), be that a kind of PET scan system output data commonly stores lattice
One of formula can convert data to more intuitive image.
Step 108, the corresponding string figure of all delay events that image reconstruction processor is acquired according to pet detector, to symbol
Random coincidence event in conjunction event is corrected.
Specifically, due to using the random coincidence event of pet detector acquisition when delay window method and using delay window method
When pet detector acquisition random coincidence event, pet detector detection count in distribution it is substantially coincident, so
The string figure for the delay event that pet detector acquires in delay time window, it is believed that being pet detector adopts meeting in time window
The estimation for meeting random coincidence event in event of collection, i.e., the string figure of all delay events is to meet random coincidence event in event
Estimation.It obtains meeting in event after the estimation of random coincidence event, the random coincidence event met in event can be removed,
Realize the correction to random coincidence event in event is met.
In the bearing calibration of above-mentioned random coincidence event, meet event and delay event for collected in pet detector
When being transferred to image reconstruction processor, by carrying out packet loss operation to delay event, reduces delay event and communication bandwidth is accounted for
With rate, and avoid the problem that delay event excessively occupies communication bandwidth and causes to meet event loss, image reconstruction processor exists
After receiving the section retards event for meeting and obtaining after event and packet loss, section retards event is carried out by deep learning network
Repairing, according to the section retards event after repairing, the random coincidence event met in event is corrected, thus improve with
Machine meets the effect of event correction, further improves PET image reconstruction effect.
In some embodiments, in training deep learning network, training dataset is obtained, it includes using that training data, which is concentrated,
Section retards event and the corresponding packet loss of delay event after trained delay event, the delay event packet loss are
Convenient for description, the section retards event after the delay event for being used for training, the delay event packet loss is referred to as delay instruction
Practice event and section retards training event.The packet loss of the string figure sum of the delay event of training will be used for as input, will be used for
The string figure of trained section retards event is trained deep learning network by Training mode as label, from
And deep learning network is improved to the repair efficiency of section retards event.
In some embodiments, packet loss operation is carried out to delay event according to default packet loss, wherein default packet loss can
It is arranged by system, it can also be by user setting.
In some embodiments, when carrying out packet loss operation to delay event, according to specified time interval to delay event
Packet loss operation is carried out, to improve PET without repairing to delay event in the time interval for not carrying out packet loss operation
Image reconstruction efficiency.As illustratively, when packet loss is divided into 100ms between 50% and specified time, can be specified every one
Time interval carries out packet loss operation to delay event, if currently assigned time interval, which outputs, meets event and delay event,
It is only exported at next specified time interval and meets event.
In some embodiments, when carrying out packet loss operation to delay event, the current counting rate of pet detector is obtained,
Dynamic adjustment is carried out according to packet loss of the counting rate to delay event, such as when the count rate is higher, higher packet loss is set
Rate carries out packet loss operation to delay event according still further to the packet loss of dynamic adjustment, so that the packet loss effect to delay event is improved,
The stability of improve data transfer amount.
In some embodiments, after carrying out packet loss operation to delay event according to the packet loss of dynamic adjustment, computing relay
The string figure and average packet loss ratio of section retards event are input in deep learning network by the average packet loss ratio of event, and generation is repaired
The string figure of section retards event after benefit, to improve the repair efficiency of delay event.Further, be delayed event average packet loss ratio
Calculation formula are as follows:
Wherein, NiFor TiThe counting of output par, c delay event, D in periodiFor TiIn period
Packet loss, m be the period number, packet loss is constant within the same period.
In some embodiments, since the string figure dimension size that different PET systems generate may be different, acquisition time is not
It is larger with the numerical fluctuations for also resulting in string figure, before the string figure of section retards event is inputted deep learning network model,
Data normalization processing is carried out to the string figure of section retards event, so that numerical value limitation of the data dimension of string figure together and in string figure
In a certain range (in the range of such as 0 to 1), to improve the repair efficiency of delay event.
In some embodiments, pet detector acquisition, which meets in time window, meets prolonging in event and delay time window
After slow event, judge whether the current count rate of pet detector is more than that preset counting rate threshold value is not required to when being less than
Packet loss operation is carried out to delay event, and all of acquisition are directly met into event and all delay events are sent to image reconstruction
Processor completes the estimation to random coincidence event, to protect as far as possible in the string figure of image reconstruction processor building delay event
Demonstrate,prove the integrality of delay event.Wherein, counting rate threshold value can be according to transmission bandwidth between pet detector and image reconstruction processor
It is configured.
It should be understood that although each step in the flow chart of Fig. 1 is successively shown according to the instruction of arrow, this
A little steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly state otherwise herein, these steps
It executes there is no the limitation of stringent sequence, these steps can execute in other order.Moreover, at least part in Fig. 1
Step may include that perhaps these sub-steps of multiple stages or stage are executed in synchronization to multiple sub-steps
It completes, but can execute at different times, the execution sequence in these sub-steps or stage, which is also not necessarily, successively to be carried out,
But it can be executed in turn or alternately at least part of the sub-step or stage of other steps or other steps.
In some embodiments, as shown in Fig. 2, providing a kind of medical imaging apparatus 200, comprising: pet detector 202
With image reconstruction processor 204, in which:
Pet detector 202, for acquiring the delay event met in event and delay time window met in time window,
And packet loss operation is carried out to delay event, section retards event and packet loss in delay event are obtained, event, part will be met
The packet loss of delay event and delay event is sent to image reconstruction processor 204.
Specifically, pet detector 202 acquires the event that meets met in time window, and generates the symbol according to delay window method
The corresponding delay time window of time window is closed, the delay event in delay time window is acquired.Pet detector 202 will meet event and
When delay event is transmitted to image reconstruction processor 204, packet loss operation is carried out to delay event, obtains section retards event.
In some embodiments, pet detector 202 is carried out by different transmission channels to event and delay event is met
Transmission, individually to carry out the integrality that packet loss is operated without influencing to meet event to delay event.
Image reconstruction processor 204, for section retards event to be rearranged to string figure, and it is corresponding according to section retards event
String figure and packet loss, the corresponding string figure of delay event is obtained by preparatory trained deep learning network, according to delay thing
The corresponding string figure of part, is corrected the random coincidence event met in event.
Specifically, meet event, section retards event and the packet loss for postponing event when image reconstruction processor 204 receives
When rate, the corresponding string figure of section retards event can be constructed, section retards event and the corresponding string figure input of packet loss are trained
Deep learning network, to be repaired by deep learning network to the section retards event, obtain deep learning network it is defeated
String figure out, the string figure of the output is string figure corresponding to all delay events of the acquisition of pet detector 202, thus completion pair
Postpone the recovery of event after packet loss.
Specifically, due to using the random coincidence event of the acquisition of pet detector 202 when delay window method and using delay window
The random coincidence event that pet detector 202 acquires when method, the distribution in the detection of pet detector 202 counts are substantially coincident
, so the string figure for the delay event that pet detector 202 acquires in delay time window, it is believed that be that pet detector 202 exists
Meet the estimation for meeting random coincidence event in event acquired in time window, i.e., the string figure of all delay events is to meet event
The estimation of middle random coincidence event.Obtain meeting in event after the estimation of random coincidence event, can by meet in event with
Machine meets event removal, realizes the correction to random coincidence event in event is met.
In some embodiments, in training deep learning network, training dataset is obtained, it includes using that training data, which is concentrated,
Section retards event and the corresponding packet loss of delay event after trained delay event, the delay event packet loss are
Convenient for description, the section retards event after the delay event for being used for training, the delay event packet loss is referred to as delay instruction
Practice event and section retards training event.The packet loss of the string figure sum of the delay event of training will be used for as input, will be used for
The string figure of trained section retards event is trained deep learning network by Training mode as label, from
And deep learning network is improved to the repair efficiency of section retards event.
In some embodiments, packet loss operation is carried out to delay event according to default packet loss, wherein default packet loss can
It is arranged by system, it can also be by user setting.
In some embodiments, when carrying out packet loss operation to delay event, according to specified time interval to delay event
Packet loss operation is carried out, to improve PET without repairing to delay event in the time interval for not carrying out packet loss operation
Image reconstruction efficiency.As illustratively, when packet loss is divided into 100ms between 50% and specified time, can be specified every one
Time interval carries out packet loss operation to delay event, if currently assigned time interval, which outputs, meets event and delay event,
It is only exported at next specified time interval and meets event.
In some embodiments, when carrying out packet loss operation to delay event, the current counting of pet detector 202 is obtained
Rate carries out dynamic adjustment according to packet loss of the counting rate to delay event, such as when the count rate is higher, higher lose is arranged
Packet rate carries out packet loss operation to delay event according still further to the packet loss of dynamic adjustment, imitates to improve to the packet loss of delay event
Fruit, the stability of improve data transfer amount.
In some embodiments, after carrying out packet loss operation to delay event according to the packet loss of dynamic adjustment, computing relay
The string figure and average packet loss ratio of section retards event are input in deep learning network by the average packet loss ratio of event, and generation is repaired
The string figure of section retards event after benefit, to improve the repair efficiency of delay event.Further, be delayed event average packet loss ratio
Calculation formula are as follows:
Wherein, NiFor TiThe counting of output par, c delay event, D in periodiFor TiIn period
Packet loss, m be the period number, packet loss is constant within the same period.
In some embodiments, since the string figure dimension size that different PET systems generate may be different, acquisition time is not
It is larger with the numerical fluctuations for also resulting in string figure, before the string figure of section retards event is inputted deep learning network model,
Data normalization processing is carried out to the string figure of section retards event, so that numerical value limitation of the data dimension of string figure together and in string figure
In a certain range (in the range of such as 0 to 1), to improve the repair efficiency of delay event.
In some embodiments, the acquisition of pet detector 202 meets meeting in event and delay time window in time window
After delay event, judge whether the current count rate of pet detector 202 is more than preset counting rate threshold value, when being less than,
It does not need to carry out delay event packet loss operation, all of acquisition is directly met into event and all delay events are sent to image
Reconstruction processor 204 completes the estimation to random coincidence event in the string figure of 204 device of image reconstruction process building delay event,
To guarantee the integrality of delay event as far as possible.Wherein, counting rate threshold value can be according to pet detector 202 and image reconstruction process
Transmission bandwidth is configured between device 204.
In some embodiments, as shown in figure 3, providing a kind of means for correcting 300 of random coincidence event, comprising: thing
Part acquisition module 302, event packet loss module 304, delay repairing module 306 and random correction module 308, in which:
Event acquisition module 302, for acquiring the delay thing met in event and delay time window met in time window
Part.
Specifically, acquisition meets the event that meets in time window, and generates this according to delay window method and meet time window correspondence
Delay time window, acquire delay time window in delay event.
Event packet loss module 304 carries out packet loss operation for collected delay event, obtains the part in delay event
Delay event and packet loss, output meet event, section retards event and the packet loss for postponing event.
Specifically, when output meets event and delay event, packet loss operation is carried out to delay event, obtains section retards
Event.
In some embodiments, by different transmission channels to event is met and delay event is transmitted, with independent
The integrality that packet loss is operated without influencing to meet event is carried out on delay event.
Delay repairing module 306, for section retards event to be rearranged to string figure, and it is corresponding according to section retards event
String figure and packet loss, the corresponding string figure of all delay events acquired by preparatory trained deep learning network.
Specifically, after meeting event, section retards event and postponing the packet loss output of event, section retards thing is constructed
Section retards event and the corresponding string figure of packet loss are inputted trained deep learning network, to pass through by the corresponding string figure of part
Deep learning network repairs the section retards event, obtains the string figure of deep learning network output, the string figure of the output
String figure corresponding to all delay events for pet detector acquisition, to complete the recovery to event is postponed after packet loss.
Random correction module 308, for the corresponding string figure of all delay events according to acquisition, to meet in event with
Machine meets event and is corrected.
The random coincidence event that is acquired when specifically, due to using delay window method and while not using delay window method acquire with
Machine meets event, and the distribution in detection counts is substantially coincident, so the delay event acquired in delay time window
String figure, it is believed that it is to meet the estimation for meeting random coincidence event in event acquired in time window, i.e., all delay events
String figure be the estimation for meeting random coincidence event in event.It obtains meeting in event after the estimation of random coincidence event, it can be with
The random coincidence event met in event is removed, realizes the correction to random coincidence event in event is met.
The specific restriction of means for correcting about random coincidence event may refer to above for random coincidence event
The restriction of bearing calibration, details are not described herein.Modules in the means for correcting of above-mentioned random coincidence event can whole or portion
Divide and is realized by software, hardware and combinations thereof.Above-mentioned each module can be embedded in the form of hardware or independently of computer equipment
In processor in, can also be stored in a software form in the memory in computer equipment, in order to processor calling hold
The corresponding operation of the above modules of row.
In some embodiments, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 4.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment meets the data such as event, delay event for storing.The network interface of the computer equipment be used for it is outer
The terminal in portion passes through network connection communication.A kind of school of random coincidence event is realized when the computer program is executed by processor
Correction method.
It will be understood by those skilled in the art that structure shown in Fig. 4, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In some embodiments, a kind of computer equipment, including memory and processor are provided, is stored in memory
Computer program, the processor perform the steps of when executing computer program
Pet detector acquisition meets the delay event met in event and delay time window in time window；
Pet detector carries out packet loss operation to collected delay event, obtains the section retards event in delay event
And packet loss, the packet loss for meeting event, section retards event and delay event is sent to image reconstruction processor；
Section retards event is rearranged to string figure by image reconstruction processor, and according to the corresponding string figure of section retards event and
Packet loss obtains the corresponding string figure of all delay events of pet detector acquisition by preparatory trained deep learning network；
The corresponding string figure of all delay events that image reconstruction processor is acquired according to pet detector, to meeting in event
Random coincidence event be corrected.
In some embodiments, it is also performed the steps of when processor executes computer program
Packet loss operation is carried out to delay event according to specified time interval.
In some embodiments, it is also performed the steps of when processor executes computer program
Packet loss operation is carried out to collected delay event according to default packet loss.
In some embodiments, it is also performed the steps of when processor executes computer program
Data normalization processing is carried out to the corresponding string figure of section retards event, section retards event is corresponding by treated
String figure and packet loss be input in deep learning network.
In some embodiments, it is also performed the steps of when processor executes computer program
The current counting rate of pet detector is obtained, dynamic adjustment is carried out according to packet loss of the counting rate to delay event；
Packet loss operation is carried out to delay event according to the packet loss of dynamic adjustment.
In some embodiments, it is also performed the steps of when processor executes computer program
According to the average packet loss ratio of the packet loss computing relay event of dynamic adjustment；
The corresponding string figure of section retards event and average packet loss ratio are input in deep learning network, PET detection is generated
The corresponding string figure of all delay events of device acquisition.
In some embodiments, it is also performed the steps of when processor executes computer program
Whether the current count rate for judging pet detector is more than preset counting rate threshold value；
When the current count rate of pet detector is less than counting rate threshold value, event will be met and delay event is sent to
Image reconstruction processor；
All delay events that image reconstruction processor acquires pet detector are rearranged to string figure, according to delay event pair
The string figure answered is corrected the random coincidence event met in event.
In some embodiments, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program performs the steps of when being executed by processor
Pet detector acquisition meets the delay event met in event and delay time window in time window；
Pet detector carries out packet loss operation to collected delay event, obtains the section retards event in delay event
And packet loss, the packet loss for meeting event, section retards event and delay event is sent to image reconstruction processor；
Section retards event is rearranged to string figure by image reconstruction processor, and according to the corresponding string figure of section retards event and
Packet loss obtains the corresponding string figure of all delay events of pet detector acquisition by preparatory trained deep learning network；
The corresponding string figure of all delay events that image reconstruction processor is acquired according to pet detector, to meeting in event
Random coincidence event be corrected.
In some embodiments, it is also performed the steps of when computer program is executed by processor
According to the packet loss of delay event, packet loss behaviour is carried out to delay event every preset quantity specified time interval
Make.
In some embodiments, it is also performed the steps of when computer program is executed by processor
Packet loss operation is carried out to delay event according to specified time interval.
In some embodiments, it is also performed the steps of when computer program is executed by processor
Packet loss operation is carried out to collected delay event according to default packet loss.
In some embodiments, it is also performed the steps of when computer program is executed by processor
Data normalization processing is carried out to the corresponding string figure of section retards event, section retards event is corresponding by treated
String figure and packet loss be input in deep learning network.
In some embodiments, it is also performed the steps of when computer program is executed by processor
The current counting rate of pet detector is obtained, dynamic adjustment is carried out according to packet loss of the counting rate to delay event；
Packet loss operation is carried out to delay event according to the packet loss of dynamic adjustment.
In some embodiments, it is also performed the steps of when computer program is executed by processor
According to the average packet loss ratio of the packet loss computing relay event of dynamic adjustment；
The corresponding string figure of section retards event and average packet loss ratio are input in deep learning network, PET detection is generated
The corresponding string figure of all delay events of device acquisition.
In some embodiments, it is also performed the steps of when computer program is executed by processor
Whether the current count rate for judging pet detector is more than preset counting rate threshold value；
When the current count rate of pet detector is less than counting rate threshold value, event will be met and delay event is sent to
Image reconstruction processor；
All delay events that image reconstruction processor acquires pet detector are rearranged to string figure, according to delay event pair
The string figure answered is corrected the random coincidence event met in event.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The several embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection of the application
Range.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of bearing calibration of random coincidence event, which is characterized in that the described method includes:
Pet detector acquisition meets the delay event met in event and delay time window in time window；
The pet detector carries out packet loss operation to the collected delay event, obtains the part in the delay event
The packet loss for meeting event, the section retards event and the delay event is sent to by delay event and packet loss
Image reconstruction processor；
The section retards event is rearranged to string figure by described image reconstruction processor, and corresponding according to the section retards event
String figure and the packet loss, all prolonging of the pet detector acquisition is obtained by preparatory trained deep learning network
The corresponding string figure of slow event；
The corresponding string figure of all delay events that described image reconstruction processor is acquired according to the pet detector, to the symbol
Random coincidence event in conjunction event is corrected.
2. the method according to claim 1, wherein the pet detector is to the collected delay event
The step of carrying out packet loss operation, comprising:
Packet loss operation is carried out to the delay event according to specified time interval.
3. the method according to claim 1, wherein the pet detector is to the collected delay event
The step of carrying out packet loss operation, comprising:
Packet loss operation is carried out to the collected delay event according to default packet loss.
4. the method according to claim 1, wherein according to the corresponding string figure of the section retards event and described
Packet loss, all delay events for obtaining the pet detector acquisition by preparatory trained deep learning network are corresponding
The step of string figure, further includes:
Data normalization processing is carried out to the corresponding string figure of the section retards event, it will treated the section retards event
Corresponding string figure and the packet loss are input in the deep learning network.
5. the method according to claim 1, wherein the pet detector is to the collected delay event
The step of carrying out packet loss operation, further includes:
The current counting rate of the pet detector is obtained, is moved according to packet loss of the counting rate to the delay event
State adjustment；
Packet loss operation is carried out to the delay event according to the packet loss of dynamic adjustment.
6. according to the method described in claim 5, it is characterized in that, according to the corresponding string figure of the section retards event and described
Packet loss, all delay events for obtaining the pet detector acquisition by preparatory trained deep learning network are corresponding
The step of string figure, further includes:
The average packet loss ratio of the delay event is calculated according to the packet loss that the dynamic adjusts；
The corresponding string figure of the section retards event and the average packet loss ratio are input in the deep learning network, generated
The corresponding string figure of all delay events of the pet detector acquisition.
7. the method according to claim 1, wherein pet detector acquisition, which meets in time window, meets event
After the step of delay event in delay time window, the method also includes:
Whether the current count rate for judging the pet detector is more than preset counting rate threshold value；
When the current count rate of the pet detector is less than the counting rate threshold value, meets event by described and described prolong
Slow event is sent to image reconstruction processor；
The delay event is rearranged to string figure by described image reconstruction processor, right according to the corresponding string figure of the delay event
The random coincidence event met in event is corrected.
8. a kind of medical imaging apparatus, which is characterized in that the medical imaging apparatus includes:
Pet detector, for acquiring the delay event met in event and delay time window met in time window, and to described
Delay event carry out packet loss operation, obtain the section retards event and packet loss in the delay event, by it is described meet event,
The packet loss of the section retards event and the delay event is sent to image reconstruction processor；And
Described image reconstruction processor, for the section retards event to be rearranged to string figure, and according to the section retards thing
The corresponding string figure of part and the packet loss obtain the corresponding string of the delay event by preparatory trained deep learning network
Figure, according to the corresponding string figure of the delay event, is corrected the random coincidence event met in event.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 7 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 7 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811224562.9A CN109124672B (en) | 2018-10-19 | 2018-10-19 | Method, device, equipment and storage medium for correcting random coincidence events |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811224562.9A CN109124672B (en) | 2018-10-19 | 2018-10-19 | Method, device, equipment and storage medium for correcting random coincidence events |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109124672A true CN109124672A (en) | 2019-01-04 |
| CN109124672B CN109124672B (en) | 2021-08-17 |
Family
ID=64808694
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811224562.9A Active CN109124672B (en) | 2018-10-19 | 2018-10-19 | Method, device, equipment and storage medium for correcting random coincidence events |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109124672B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109658473A (en) * | 2019-01-24 | 2019-04-19 | 上海联影医疗科技有限公司 | Actively reduce data link carrying method, device and computer equipment |
| CN110415311A (en) * | 2019-07-29 | 2019-11-05 | 上海联影医疗科技有限公司 | PET image reconstruction method, system, readable storage medium storing program for executing and equipment |
| CN112817035A (en) * | 2021-01-28 | 2021-05-18 | 上海联影医疗科技股份有限公司 | Data compensation method and device, computer equipment and storage medium |
| CN113393547A (en) * | 2021-05-25 | 2021-09-14 | 上海联影医疗科技股份有限公司 | PET (positron emission tomography) coincidence data volume control method, device, equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050253075A1 (en) * | 2004-05-14 | 2005-11-17 | Cti Pet Systems, Inc. | Nearest-neighbor rebinning in clinical PET using on-line three dimensional LOR-to-bin mapping |
| CN103417235A (en) * | 2013-07-31 | 2013-12-04 | 沈阳东软医疗系统有限公司 | Random noise correction method and device |
| CN104335247A (en) * | 2012-05-21 | 2015-02-04 | 皇家飞利浦有限公司 | Fast scatter estimation in PET reconstruction. |
| CN106233336A (en) * | 2014-04-23 | 2016-12-14 | 皇家飞利浦有限公司 | PET accidental coincidence correction |
| CN107137101A (en) * | 2017-04-24 | 2017-09-08 | 沈阳东软医疗系统有限公司 | A kind of time calibrating method and device |
-
2018
- 2018-10-19 CN CN201811224562.9A patent/CN109124672B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050253075A1 (en) * | 2004-05-14 | 2005-11-17 | Cti Pet Systems, Inc. | Nearest-neighbor rebinning in clinical PET using on-line three dimensional LOR-to-bin mapping |
| CN104335247A (en) * | 2012-05-21 | 2015-02-04 | 皇家飞利浦有限公司 | Fast scatter estimation in PET reconstruction. |
| CN103417235A (en) * | 2013-07-31 | 2013-12-04 | 沈阳东软医疗系统有限公司 | Random noise correction method and device |
| CN106233336A (en) * | 2014-04-23 | 2016-12-14 | 皇家飞利浦有限公司 | PET accidental coincidence correction |
| CN107137101A (en) * | 2017-04-24 | 2017-09-08 | 沈阳东软医疗系统有限公司 | A kind of time calibrating method and device |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109658473A (en) * | 2019-01-24 | 2019-04-19 | 上海联影医疗科技有限公司 | Actively reduce data link carrying method, device and computer equipment |
| CN109658473B (en) * | 2019-01-24 | 2023-08-01 | 上海联影医疗科技股份有限公司 | Method, device and computer equipment for actively reducing load of data link |
| CN110415311A (en) * | 2019-07-29 | 2019-11-05 | 上海联影医疗科技有限公司 | PET image reconstruction method, system, readable storage medium storing program for executing and equipment |
| CN110415311B (en) * | 2019-07-29 | 2024-04-16 | 上海联影医疗科技股份有限公司 | PET image reconstruction method, system, readable storage medium and apparatus |
| CN112817035A (en) * | 2021-01-28 | 2021-05-18 | 上海联影医疗科技股份有限公司 | Data compensation method and device, computer equipment and storage medium |
| WO2022161478A1 (en) * | 2021-01-28 | 2022-08-04 | Shanghai United Imaging Healthcare Co., Ltd. | System and method for pet data compensation |
| CN113393547A (en) * | 2021-05-25 | 2021-09-14 | 上海联影医疗科技股份有限公司 | PET (positron emission tomography) coincidence data volume control method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109124672B (en) | 2021-08-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109124672A (en) | Bearing calibration, device, equipment and the storage medium of random coincidence event | |
| DE112011101470T5 (en) | Adaptive wireless sensor network and method for routing data in a wireless sensor network | |
| CN109658473B (en) | Method, device and computer equipment for actively reducing load of data link | |
| CN106821402B (en) | Method and device for constructing PET image | |
| US20190012812A1 (en) | Calibrating Time In PET Device | |
| CN110327067A (en) | Image rebuilding method, device, terminal device and PET system | |
| CN109492253B (en) | Method and device for evaluating radiation damage reliability of semiconductor device | |
| CN106388841A (en) | An image reconstruction method, device and apparatus | |
| CN105763818B (en) | The flat panel detector with image correction function based on FPGA | |
| CN103685060A (en) | Data packet sending method and data packet sending device | |
| US20240122567A1 (en) | Pet-data correction method and device, computer apparatus, and pet-image reconstruction method | |
| CN106372026A (en) | Link detection method and receiving device | |
| CN106821406A (en) | Medical image system | |
| CN113762197A (en) | Transformer substation fire detection method and device based on terminal power business edge calculation | |
| CN110288008A (en) | Look-up table method for correcting image, device, computer equipment and storage medium | |
| CN110400361B (en) | Method, device and computer equipment for subset division and image reconstruction | |
| CN108132592A (en) | A kind of time-to-digital conversion apparatus, detector, method and medium | |
| CN104598356A (en) | Event ordering method and device | |
| CN110364211B (en) | Method, device and equipment for reducing erasing interference time of nonvolatile memory | |
| CN108703769B (en) | Method, device and system for correcting TOF data and computer-readable storage medium | |
| Leuschner et al. | Towards automated configuration of systems with non-functional constraints | |
| CN109426595A (en) | A kind of analysis system, method and the application of FPGA log | |
| CN110432923A (en) | Image rebuilding method, device, computer equipment and storage medium | |
| CN112102426B (en) | Background coincidence event judging and selecting method, device and equipment and readable storage medium | |
| Tsoumpas et al. | Convergence properties of algorithms for direct parametric estimation of linear models in dynamic PET |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 201807 2258 Chengbei Road, Jiading District, Shanghai Applicant after: Shanghai Lianying Medical Technology Co., Ltd Address before: 201807 2258 Chengbei Road, Jiading District, Shanghai Applicant before: SHANGHAI UNITED IMAGING HEALTHCARE Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |