WO1998023973A1 - Correction de la diffusion compton en temps reel - Google Patents

Correction de la diffusion compton en temps reel Download PDF

Info

Publication number
WO1998023973A1
WO1998023973A1 PCT/IL1996/000162 IL9600162W WO9823973A1 WO 1998023973 A1 WO1998023973 A1 WO 1998023973A1 IL 9600162 W IL9600162 W IL 9600162W WO 9823973 A1 WO9823973 A1 WO 9823973A1
Authority
WO
WIPO (PCT)
Prior art keywords
event
energy
radiation
events
filter function
Prior art date
Application number
PCT/IL1996/000162
Other languages
English (en)
Inventor
Gideon Berlad
Dov Maor
Original Assignee
Ge Medical Systems Israel, Ltd.
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Ge Medical Systems Israel, Ltd. filed Critical Ge Medical Systems Israel, Ltd.
Priority to PCT/IL1996/000162 priority Critical patent/WO1998023973A1/fr
Priority to JP52643398A priority patent/JP2001516446A/ja
Priority to US09/308,376 priority patent/US6369389B1/en
Publication of WO1998023973A1 publication Critical patent/WO1998023973A1/fr

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
    • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
    • G01T1/16—Measuring radiation intensity
    • G01T1/161—Applications in the field of nuclear medicine, e.g. in vivo counting
    • G01T1/164—Scintigraphy
    • G01T1/1641—Static instruments for imaging the distribution of radioactivity in one or two dimensions using one or several scintillating elements; Radio-isotope cameras
    • G01T1/1648—Ancillary equipment for scintillation cameras, e.g. reference markers, devices for removing motion artifacts, calibration devices
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
    • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
    • G01T1/16—Measuring radiation intensity
    • G01T1/161—Applications in the field of nuclear medicine, e.g. in vivo counting
    • G01T1/164—Scintigraphy
    • G01T1/1641—Static instruments for imaging the distribution of radioactivity in one or two dimensions using one or several scintillating elements; Radio-isotope cameras
    • G01T1/1642—Static instruments for imaging the distribution of radioactivity in one or two dimensions using one or several scintillating elements; Radio-isotope cameras using a scintillation crystal and position sensing photodetector arrays, e.g. ANGER cameras
    • 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
    • A61B6/42—Arrangements for detecting radiation specially adapted for radiation diagnosis
    • A61B6/4208—Arrangements for detecting radiation specially adapted for radiation diagnosis characterised by using a particular type of detector
    • A61B6/4258—Arrangements for detecting radiation specially adapted for radiation diagnosis characterised by using a particular type of detector for detecting non x-ray radiation, e.g. gamma radiation

Definitions

  • Fig 1 shows a block diagram of a generalized prior art nuclear medicine system 550 used to image a patient 504 System 550 is used to generate images of radio-nuclide concentrations in patient 504
  • a radio-pharmaceutical which usually forms at least a concentration 506 in portions of patient's 504 anatomy
  • Patient 504 is placed in an examination area (not shown) so that a scintillation detector crystal 500 can detect gamma radiation emitted by radio-pharmaceuticals in concentration 506
  • gamma rays emitted by radioactive materials are treated as a particle phenomenon
  • Each measured photon corresponds to one radiation event and the number of radiation events from a region reflects the concentration of the radio-active material in that region
  • the energy of the events indicates whether they have traveled directly from concentration 506 or have their origin in a different region of the anatomy and have been scattered
  • nuclear medicine system design emphasizes filtering real radiation events from scattered events (whose origin is unknown) Due to the weak interaction between gamma-rays and matter and the desire to use low dosages of radioactivity, gamma-rays are not captured on film or with detectors such as used in X-ray CT systems
  • Gamma radiation emitted by the radio- pharmaceutical interacts with detector crystal 500 to produce minuscule flashes of light Each radiation event generates one light flash in detector 500
  • Several of a plurality of photomultipliers 502 detect this flash of light and generate an electrical current responsive to the intensity of light sensed by the individual photomultiplier The contributions of photomultipliers
  • shaper/delayer 512 is triggered only if the radiation event has a total energy which is within a specific wide energy window Otherwise, the measured event is probably an uninteresting scattering event and is discarded.
  • the outputs of all of amplifiers 508 are summed by an adder 510
  • the sum calculated by adder 510 is used by a gating unit 514 to selectably trigger shaper/delayer 512 responsive to the sum If the sum is within a preset range of values, gating unit 514 triggers shaper/delayer 512 to process the radiation event It should be noted that the total energy of the event is approximately determined at this stage Using the delay, full scale integration eventually starts only if the approximate energy falls within predefined limits
  • An integrator 516 integrates the signal produced by shaper/delayer 512 to find the total energy associated with the radiation event detected by one specific photomultiplier 502
  • An important result of the integration is noise reduction As in many measurement systems, even when there is no event being measured, there is a parasitic DC level, usually referred to as a base-line voltage This base-line voltage is typically subtracted from the signal before integration Otherwise, the integrated signal contains a large (unknown) contribution from the base-line This process is called base line restoration
  • the individual detector circuits are connected to a single event processing unit 519
  • a sequencer 517 multiplexes the results from all of integrators 516 and passes them serially to event processing unit 519
  • analog signal is converted to a digital signal after integration
  • Conversion of analog signals to digital signals is problematic for the short pulse durations typical of nuclear medicine imaging
  • analog to digital converters tend to (a) have relatively low resolutions
  • U S Patent No 5,371,362 discloses a base line measurement and correction system
  • the output signal of each photomultiplier is sampled by an analog-to-digital (A/D) converter and analyzed to determine the values of the base-line voltage between radiation events
  • the determined base line voltage is subtracted from the sampled signals prior to integration to yield base-line corrected signals
  • Also disclosed is the addition of a sliding scale voltage to the photomultiplier output signal
  • a sliding scale voltage is generated by the system responsive to the amplitude of the sampled signal
  • the sliding-scale voltage is added to the signal from the photomultipliers so that its amplitude is within the linear range of the analog to digital converter
  • a sliding scale signal having a cycle time which is 64 events long is added to the analog signal before conversion
  • Each step of the sliding scale is equivalent to about one LSB (least significant bit) of the A/D converter After A/D conversion, A digital value corresponding to the sliding scale analog value
  • event processing starts with determining the X-Y position of the radiation event on detector head 500 Only strong signals are useful for this determination Thus, a selector 518 selects only those integration results which are above a threshold A normalizer 520 normalizes the selected results to make their sum a constant and a position calculator 522 uses the normalized results to perform Anger arithmetic and calculate the position of the radiation event in the plane of detector 500
  • Linearity errors are systematic errors in the position calculation by Anger arithmetic Sensitivity errors are caused by detector 500 having a position dependent sensitivity, i e , some portions of detector 500 naturally detect more events than others portions, even if all of detector 500 is receiving a uniform event flux
  • Energy errors are caused by non-detection of some of the light generated by an event, e g , light passing through the spaces between photomultiplier tubes 502
  • similar events are acquired by system 550 as having dissimilar energy levels
  • the events whose energy level is not within a position dependent narrow window are rejected by system 550
  • events are corrected only after non-events are rejected
  • Corrector 524 can be configured to perform the transformation However, all of the geometric transformations (linearity corrections and others) are usually performed as one step using a single table Thus, when the desired transformation changes, the geometric transformation table in corrector 524 must be recalculated, which is time consuming
  • the transformations can be performed on the final image
  • the quality of the transformed image is lower than the quality of the original image, due to aliasing effects.
  • Image generation by an image processor 526 completes the processing of radiation events so that a completed image can be displayed on a display 528
  • a digital nuclear medicine system includes, a detector unit which generates an analog electrical signal responsive to an impinging radiation event The analog signal is converted to a digital signal before any steps of
  • a resolution enhancement signal is added to the analog signal before converting the analog signal to the digital signal
  • the resolution enhancement signal is a time varying cyclical or pseudo-random signal which cycles several times during the detection of a single event and has a precision higher than the resolution of the analog to digital converter
  • the resolution enhancement signal has an amplitude on the order of the resolution of the analog to digital conversion
  • the sum of the resolution enhancement signal over the integration time is zero
  • the resolution enhancement signal does not add to the integrated signal and, as a consequence, does not need to be corrected for before integration
  • the addition of the resolution enhancement signal is corrected for by subtracting a digital equivalent of the integral of the resolution enhancement signal from the digital signal before restoring the baseline
  • the resolution enhancement signal is preferably synchronized to the analog to digital conversion process
  • a second limitation of analog to digital converters is that they are not uniformly linear over their acquisition range These non-uniformities can be corrected by mapping the response of the converter and correcting some of the errors after acquisition However, even if a converter is mapped to correct for these non-uniformities, when the converter is replaced it must be mapped again In addition, the linearity of analog-to-digital converters changes over time
  • a linearity enhancement signal is added to the analog signal before conversion thereof to the digital signal
  • the linearity enhancement signal is added in addition to, or alternatively to, the resolution enhancement signal
  • the linearity enhancement signal cycles slowly, such that it approximates a constant signal during the integration time of a single event
  • the amplitude of the linearity enhancement signal is preferably approximately 5% of the range of the analog to digital converter
  • the amplitude of the linearity enhancement signal is preferably higher than the resolution of the analog to digital converter, i e , it is many resolution elements in amplitude
  • the digital signal that has its base line restored by the base line restorator has a higher precision than the digital signal generated by the analog to digital converter It should also be appreciated that since the linearity enhancement signal varies slowly, the base line restorator treats it as a DC signal so that the base line restorator can generally correct for the linearity enhancement signal without special circuitry and without receiving the linearity enhancement signal as an input Optionally, a digital equivalent of the linearity enhancement signal is subtracted from the digital signal before base line restoration
  • errors in event detection are corrected, for example, errors due to non-uniform sensitivity of the detector
  • other positioning errors in the positioning of the event are corrected, for example, errors caused by mechanical misalignment
  • (g) events are assigned to a location in an image plane Generally, when more than one of (b), (e) and (f) are performed, they are usually performed using a single transformation map Alternatively, (e) or (f) are performed on the image plane after
  • (a) - (f) are performed as separate, independent steps on each individual event
  • an event is assigned to the image plane only after all the desired transformations and corrections have been applied to it
  • each event can have different, non-constant, possibly time based, corrections and transformations applied to it
  • the precision of (a) - (e) is as high as that of the calculation system, and is not limited by the image plane resolution
  • a time-based correction in accordance with a preferred embodiment of the invention, corrects for distortions caused by some radio-pharmaceuticals deteriorating rapidly during a nuclear medicine session Radiation events which occur later in the session are given higher weights to compensate for this deterioration
  • Another preferred non-constant correction is a correction for mechanical misalignment
  • detector assemblies are rotated around a patient
  • the detector assemblies sag as a result of their weight
  • the center of rotation for the assemblies is not always exactly at the center of the projection plane
  • the error in the position of the event caused by sagging of the camera detectors is corrected by applying an angle based geometric transformation to each event
  • the center of rotation is also corrected by applying a second geometric transformation to the event
  • Yet another preferred non-constant correction is a velocity correction for linear scanning, where a detector scans along the patients body
  • velocity changes are corrected for by giving each event a position dependent weight, depending on the velocity of the detector at the time the event is acquired
  • a common geometrical transformation is converting a fan beam image to a parallel beam image to optimally utilize both detector and image plane area
  • a small portion of the patients body is imaged using a fan-beam collimator
  • the radiation events are repositioned, before being assigned to the image plane, to simulate the use of a parallel beam collimator
  • a high resolution fan-beam acquired image is displayed without the typical distortions caused by the fan beam collimator
  • each event is provided with a weight
  • several type of irregularities are correctable by varying the weighting of each event Sensitivity correction is achieved in a preferred embodiment of the present invention by first mapping the sensitivity uniformity of the detector and, during acquisition, giving each event a (fractional) weight which is dependent on the event's position in the detector
  • Compton scattering artifacts are reduced in real-time on an event-by-event basis
  • Each event is assigned using an accumulation matrix to a central pixel with a first (energy dependent) weight and to a surrounding region with a second (energy dependent) weight
  • the ratio between the first and second weights is also energy dependent
  • an event counter counts the number of actual events A higher weight is attached to an event during a period in which the event rate is high
  • events occurring during a short frame time are given a higher weight than events occurring during long frame times
  • a method of reducing artifacts caused by unwanted photons having a known energy distribution in a radiation camera which images radiation from primary radiation events including
  • the filter function has a first portion having a first weight and a second portion having a second weight and the ratio between the first and second weights is energy dependent Alternatively or additionally, the filter function is dependent on a predetermined energy dependent scatter coefficient of the unwanted photons in the radiation camera
  • the filter function has a continuous dependence on the energy of the event Additionally or alternatively, the filter function is predetermined Additionally or alternatively, the filter function is dependent on the energy of the primary radiation event
  • the primary radiation events are generated by one of at least two different radioactive elements introduced into a patient
  • unwanted photons include Compton scattered photons of a primary event
  • the unwanted photons include x-rays generated by the interaction of a primary event with the radiation camera
  • a nuclear medicine system which images radiation from primary radiation events and which detects unwanted photons having a known energy distribution, including an event detector which detects a radiation event, an event energy and location detector which determines the energy and location of the event, and an unwanted-photon corrector which receives an event energy and location and convolutes the location with an energy dependent filter function, where the filter function is based on a predetermined energy distribution of unwanted photons and on a predetermined system point spread function
  • a nuclear medicine system which images radiation from primary radiation events and which detects unwanted photons having a known energy distribution, including an event detector which detects a radiation event, an event energy and location detector which determines the energy and location of the event, and an unwanted-photon corrector which receives an event energy and location and convolutes the location with an energy dependent filter function, which filter function is a spatial filter and where the filter function at one element is different in more than magnitude than the filter function at a different energy associated with a primary event having the same energy
  • the filter function includes a first weight and a second weight and the ratio between the first and second weights is energy dependent
  • different filter functions are used for different primary event types
  • the unwanted photons include Compton scattered photons of a primary event
  • the unwanted photons include x-rays generated by the interaction of a primary event with the radiation camera
  • FIG. 1 is a block diagram of a prior art nuclear medicine system
  • Fig 2A is a graph showing a response of a photomultiplier tube to a series of radiation events
  • Fig. 2B shows a prior art wave form which is added to the photomultiplier response
  • Fig. 2C-2D are graphs showing various wave forms which are added to the photomultiplier response in various preferred embodiments of the present invention
  • Fig. 3 is a general block diagram of a nuclear medicine imaging system according to a preferred embodiment of the invention.
  • Fig 4 is a block diagram of a detector circuit of the system shown in Fig. 3
  • Fig. 5 is a block diagram of a processor of the system shown in Fig. 3
  • Fig. 6 is a block diagram of gating circuit of the system shown in Fig 4
  • Fig. 7 illustrates, in an exaggerated manner, the effect of sagging on a detector crystal imaging a patient
  • Fig 3 is a general schematic of a nuclear medicine system 61 according to a preferred embodiment of the invention.
  • System 61 generally comprises an examination area 44, a detector crystal 41 and a plurality of photomultiplier (PM) tubes 14 When a gamma-ray source is placed in examination area 44, gamma-rays created by a radiation event interact with detector 41 to generate weak scintillations These scintillations are amplified by PM tubes 14, which also convert the scintillations into electrical signals having an amplitude related to the energy of the interacting gamma-ray Each of PM tubes 14 is connected to a detector circuit 34, described more fully below The outputs from circuits 34 are passed to a processor 60 which determines the position of each radiation event from the signals generated in PM tubes 14 and combines these events to form an image This resulting image is displayed on a display 92
  • System 61 is generally controlled by a controller 100 which generates controlling and timing signals In a typical process of nuclear medicine image acquisition, a patient 40 is injected with
  • Fig 4 show a block diagram of detector circuit 34 according to a preferred embodiment of the present invention
  • the gain of each PM 14 is controlled by a gain controller 12, since each PM typically amplifies incident light by a different factor.
  • gain controller 12 is controlled by controller 100 to compensate for known deviations in the amplification of individual PMs 14
  • the pulse signal is preferably further amplified by a pre-amplifier 16
  • pre-amplifier 16 performs a small amount of smoothing to reduce noise
  • pre-amplifier 16 has a bandwidth of 5-8 MHz
  • the smoothed pulse signal is digitized by an A/D converter 20 after processing by an adder 18, as described below It should be noted that A D converter 20 acquires a number of data samples during each radiation event, so that the pulse signal can be reproduced from the digitized data
  • A/D converter 20 samples at a rate of 20 MHz
  • the resolution of A/D converter 20 is preferably at least 8 bits, and most preferably 10 bits or more.
  • a resolution enhancement signal is added to the analog signal using adder 18 Fig 2 A shows a series of analog signals generated by a PM in response to radiation events Fig. 2C, which shares the same time scale as Fig. 2A, shows a preferred resolution enhancement signal.
  • the effect of the resolution enhancement signal is best explained by the example of the acquisition of a constant signal
  • one bit step of A/D converter 20 is equal to 8 millivolts
  • the digitized signal has the value 1, even though the real equivalent digital value is 1 875, since only the "whole" portion of the signal is digitized
  • the constant signal is dithered by a periodic signal with an amplitude of 8 millivolts (one bit)
  • the resulting analog signal is between 7 and 23 millivolts
  • the digitized signal is either 1 or 2
  • the average digital signal will approach 1 375, since in 7/16 of the cases the signal voltage will be above 16 millivolts and in 9/16 between 8 and 16 millivolts 0 5 (half a bit) is added to the average signal to reflect the average rounding error of the A/D, i.e , the fact that all values between 8 and 15 millivolts
  • the average acquired value is 1 875
  • the additional resolution is log(N) if each digitization is performed at a different phase of the periodic dithering signal
  • the dithering periodic signal cycles approximately two times during a pulse signal, which is approximately 800 ns
  • the periodic signal is synchronized to the digitization clock, to ensure that each digitization is at a different phase of the periodic signal
  • the digitization frequency is set to be four times the frequency of the periodic signal
  • the timing of the digitization is set so that within each cycle of the periodic signal, each digitization is at a different phase of the periodic signal This timing repeats in consecutive signal cycles Alternatively, a timing scheme which repeats less often than every cycle is used
  • the resolution enhancement effect may not be achieved for a single pulse signal, especially if the frequency of the resolution enhancement signal is low compared to the pulse signal
  • the first sample may be elevated from 1 9 to 2 1, (I e , from 1 to 2) and the second sample reduced from 1 3 to 1 1 (i e , not affected)
  • the resolution enhancement signal has a higher frequency and where the amplitude of the pulse is non-constant, the accuracy of most all events is increased
  • the periodic signal is also synchronized to the integration.
  • the integration time of a pulse signal is an integral multiple of the period of the resolution enhancement signal, to ensure that the sum of contributions of the periodic signal to the integrated value is zero.
  • the digital value of the resolution enhancement signal at the sample points is preferably subtracted from the digitized signal, as described below.
  • a useful embodiment of a periodic wave form for variable integration systems is a saw tooth wave.
  • the resolution enhancement signal is controlled by controller 100, as described below.
  • the resolution enhancement signal is produced by a digital sliding scale source 10.
  • control and timing signals for digital sliding scale source 10 are generated by controller 100.
  • the amplitude of the resolution enhancement signal is preferably at least as high as a resolution step of the A D converter 20.
  • the accuracy of the resolution enhancement signal is preferably higher than the desired additional resolution of digitization.
  • A/D converters do not usually have a linear response over their entire dynamic range.
  • the linearity of A/D converters changes over time, particularly if the converter is replaced.
  • Some analog to digital conversion systems use a linearity map to correct these non- linearities.
  • using a linearity map lowers the accuracy of the digitized signal.
  • the map needs to be updated periodically, especially if the A/D converter is replaced.
  • Fig. 2D shows a linearity enhancement signal (not to scale) which is preferably added to the analog signal before digitization.
  • This linearity enhancement signal preferably has a much lower frequency than the resolution enhancement signal, typically 60-100 Hz.
  • the amplitude of the linearity enhancement signal is preferably about 5% of the total range of A/D converter 20.
  • each radiation event response is digitized in a slightly different region of the dynamic range of A/D converter 20 and non-linearities are averaged out. Because the linearity enhancement signal changes very slowly compared to the duration of a radiation event response, it is perceived as a DC signal by the rest of detector circuit 34.
  • the linearity enhancement signal is generated by a digital sliding scale source 11 and added to the analog signal by adder 18
  • the control and timing signal for each digital sliding scale source 1 1 are preferably generated by controller 100
  • each detector circuit 34 has its own digital sliding scale source 1 1 and these sources are not synchronized between different detector circuits
  • each detector circuit 34 also has its own, unsynchronized digital sliding scale source 10
  • Embodiments of the invention may employ one or both types of enhancement signals
  • a linearity enhancement signal according to the instant invention is not a baseline restoration signal, as described for example in U S Patent No 5,508,425 to Goldberg et al, the disclosure of which is incorporated herein by reference, since the value is not dependent on a preset base-line offset On the contrary, the linearity enhancement signal purposely introduces a time-varying base-line offset
  • the digital value of the resolution enhancement signal and/or the linearity enhancement signal is subtracted from the digitized signal to lower noise levels in the digitized signal
  • the subtraction is typically not needed
  • the base line of the signal is measured during times when no radiation event is detected, the accuracy of measurement of the base line is enhanced by the resolution enhancement signal
  • the linearity enhancement signal does not generally add to the accuracy of the measured base line, since it adds a constant to the real base line
  • the slowly changing linearity enhancement signal acts like part of the real base line for restoration purposes
  • a shaper-delayer 24 reshapes the digital signal, after it is sampled, to produce a new digital signal with an equivalent energy, but a shorter temporal extent In effect, the signal is compressed
  • shaper-delayer 24 is controllable by controller 100 to adjustably shape pulses in a parametric manner
  • pulses may be shaped differently as a function of event rate
  • the present invention preferably uses a digital filter to shape/delay Before integrating the digital signal to determine the total energy detected by an individual PM 14, it is important to subtract the baseline component of the signal This baseline is generally equal to the DC signal detected when no events (or scattering events) are
  • B(t) is the baseline at time t
  • dt is the sampling time
  • Signal(t) is the input to restorator 26 at time t
  • m the number of samples
  • n is the window size
  • a gating circuit 32 determines when no radiation events are occurring Digital signals acquired while no radiation event response is present are entered into the window to estimate the base line for signals acquired while a radiation event response is present An integrator 28 is activated by gating circuit 32 to integrate digital signals acquired during a radiation event response When the radiation event response is over (also determined by gating circuit 32), the integrated signal represents the total amount of energy from the radiation event detected by PM 14
  • Gating circuit 32 is used to differentiate between signals which are responses to radiation events and clutter signals such as random noise signals and scattered radiation events Typically, if the sum of the detected signals from all PMs 14 is outside a specific energy window, the signal is discarded as unusable Otherwise, it is treated as a valid event, until a further energy value determining step The sum is determined by adding the pulse signals from all of PMs 14, smoothing the resultant meta-pulse and searching for a peak value, which is compared to an energy window Preferably, the extent of the energy window is controlled by controller 100 based on the system noise levels, radio-pharmaceutical type and count rate
  • Fig 6 shows a preferred embodiment of a complete gating apparatus, including an adder 40 which sums the energy readings from all of PMs 14
  • a wide window rejecter 42 smoothes the sum (which is relatively an analog signal), and determines whether the peak of the energy sum signal is within an amplitude window. If the peak is within the window, rejecter 42 sends an activation signal to gating circuit 32, which is timed to start the integration and the shaping, so that they start at the beginning of the pulse signal and end at its end. It should be noted that the pulse signal to be integrated is delayed by the delayer portion of shaper/delayer 24, to give circuits 40, 42 and 32 time to operate.
  • adder 40 sums analog signals, since analog summation is generally more precise than digital summation, for low level signals.
  • the sum is then digitized by an A/D 44 for further use as described herein.
  • the analog signals are digitized before being summed or before being passed to wide window rejecter 42, preferably using the resolution and/or linearity enhancement signals as described above.
  • the integrated signal has a linear relationship to the distance of the radiation event from the center of PM 14.
  • this relationship is distorted.
  • a main cause of distortion is that PM 14 is more sensitive at its center than near its edges.
  • Other causes include the difference in refractive index between detector crystal 41 and PM 14 and location dependent variations in detector 41.
  • the spatial response of PM 14 is modeled by a modeler 29 to correct distortions.
  • the integrated signal is corrected to reflect the value that would have been acquired had PM 14 been optimal.
  • all the PMs 14 are modeled using a single empirical (measured) model, usually a look-up-table, preferably a one dimensional look-up table.
  • a mathematical function is used instead of a look-up-table to model the spatial response.
  • the correction model may be two dimensional.
  • each PM 14 can have a different, personalized, model associated therewith.
  • PM modeling is performed in detector circuit 34 by comparing a measured PMC(x,y) to an ideal value PMC(0,0), wherein PMC(x,y) is the response of the tube at to an event which occurs at (an unknown) location (x,y) normalized to the total energy signal from A/D 44.
  • a look-up table is used to determine d/D, the ratio between the distance of the event from the PM center and the distance between two adjacent PMs, based on the difference between PMC(x,y) and PMC(0,0). Once d/D is known, the energy for this pulse signal and all the other pulse signals corresponding to the same event can be corrected as described hereinabove.
  • the total measured energy of the event is previously determined by adder 40.
  • the uncorrected flood contrast ratio can be up to 2.0. Correcting the energy of the individual PM 14 at this point in the process by modeling the photomultiplier response greatly reduces the non-linearities in position determination using Anger arithmetic Alternatively, the energy is corrected by processor 60 immediately after its receipt of the pulse signal Once the pulse signals are integrated and, optionally distortion corrected, they are passed in sequence to processor 60 as components of a single radiation event response (a meta-pulse signal) A sequencer 30 synchronizes and sequences the signals from detector circuits 34 to processor 60
  • Fig 5 is a simplified block diagram of processor 60 which forms a nuclear medicine image based on detected radiation events Since the image is formed by events, the first step performed by processor 60 is to determine the location of the radiation event in detector crystal 41 A preferred method of event location is called Anger arithmetic and is described in U S Patent No 3,011,057 However other methods of localization are well known and useful in carrying out the invention, for example, as described in U S Patent No 5,285,072 The disclosures of both Patent documents are incorporated herein by reference
  • a selector 62 selects only those PM signals which are greater than a threshold Preferably, the height of the threshold is controlled by controller 100 responsive to the total energy (sum of all the signals from adder 40)
  • the selected signals are normalized to the sum of the signals generated by all the PMs by a normalizer 68
  • a positioner 80 calculates the position of the radiation event in detector 41, using Anger arithmetic, based on the selected signals
  • an energy corrector 70 corrects the energy of the acquired event for these distortions before windowing
  • the energy window is varied as a function of the position of the event
  • Correcting energy distortions is usually accomplished by calibrating the system to determine calibration data which is stored in a look-up table
  • the exact energy value is determined by interpolating between table values
  • different look-up tables are used depending on the energy range of the radiation event.
  • a linearity corrector 72 corrects errors in the localization of radiation events
  • a configuration map which maps the linearity errors, is used to correct event positions in real-time
  • a continuous approximation such as a spatial B-spline approximation is used to interpolate between map data points
  • Different linearity correction maps may be used depending on the energy range of the radiation event
  • a sensitivity corrector 74 applies this logic by assigning a weight to each event A high weight is attached to an event which is detected in a less sensitive, low probability area and a low weight is attached to an event which is detected in a more sensitive, high probability area Preferably, decay of the injected radio-pharmaceutical is corrected for by varying the weight given to each event
  • Energy corrector 70, linearity corrector 72 and sensitivity corrector 74 may be applied in other orders than those described above, with appropriate changes to the correction algorithms
  • a geometric transformer 76 preferably applies three types of transformations to the event First, any additional geometric distortions are preferably corrected, as described in further detail below Second, static transformations such as zooming and rotating are performed, if desired Third, if detector 41 is moving relative to patient 40, (e g , during a whole body scan) the event is transformed to its proper, time dependent position in the image plane Each transformed event is binned to a location in the image plane and individual event characteristics are lost It should be noted that the image plane might have three or more dimensions, i e , spatial location in the detector, energy, spatial location of the detector and gating/binning information if physiological binning is used A frame builder 78 transforms the events binned in the image plane into a nuclear medicine image If a three-dimensional image is acquired, for example, utilizing SPECT, frame builder 78 generates a tomographic image from the data binned in the image plane Frames created by frame builder 78 are
  • each event receives a individualized transformation based on its exact characteristics
  • the corrections are performed in the following manner
  • Compton scattering artifacts are reduced in real-time, preferably by a Compton remover 77 on an event-by-event basis
  • a preferred method of Compton scattering artifacts reduction, which is not on an event-by-event basis, is described in U S Patents Nos 5,293,195 and 5,434,414, the disclosures of which are incorporated herein by reference
  • the count in a pixel element is also dependent on the events detected in neighboring pixels
  • the size of the neighborhood (S) depends on the desired noise reduction and on available computer power Image sharpness is not substantially affected by the neighborhood size (within reasonable bounds) because the Compton energy distributions decrease slowly with the distance from the (central) pixel element The Compton distribution thus results in a point spread function having a sharp peak and low, broad edges
  • the corrected count is calculated by
  • N p SL (x,y) ⁇ A l (E)- N E (x,y) + ⁇ A s (E)- N ⁇ (x,y) (3)
  • semi-local Compton scatter correction is especially important when searching for cold-spots in a hot region.
  • the broad point spread function which is caused by the Compton scattering may hide the cold-spot by completely obliterating any contrast between the hot areas and the cold-spot.
  • the events are corrected for Compton artifacts on an event- by-event basis, rather than after the fact.
  • a further advantage of event-by-event correction is that once all the events are acquired, the image is Compton free, without the necessity of performing additional processing on the image.
  • Nf (x,y) ⁇ A (E) - N(E,x,y)dE
  • TM S E) -- where N(E,x,y) is a continuous function of E, x and y, giving the number of counts at location (x,y) having an energy E and N (E,x,y) is the number of counts in area S, not
  • w,(E) and Wg,(E) are pre-calculated based on
  • J(E), P(E) and S A 32 bit fixed point matrix, having 16 bits for the integral portion and 15 bits for the fractional portion is preferably used (+ one bit for the sign) to accumulate the events
  • Each event is entered at its location (x,y) with a weight w, (E) and values W (E) are entered at several, preferably 8 neighboring pixels
  • each event is individually convoluted with a filter having energy dependent weights
  • the weight in the center pixel is usually different from the weights in the surrounding pixels
  • only local correction is performed, using a single weight J(E) (zero dimension convolution)
  • the size and shape of area S and the weights Wi and wg may be varied In particular, if a larger areas S is used, the weight assigned to distant (from the center) pixels may be different (lower) then the weights assigned to pixels near the central pixels
  • Compton scatter correction can also be used to reduce other unwanted radiation events whose energy-dependent distribution is known, such as lead X-ray events which are caused by the interaction
  • an event counter is used to count the approximate number of real events If a lower number of radiation events is processed by system 61, than the actual number of events, due to count rate limitations, the weight of processed events is increased to compensate
  • a parallel beam image is simulated using a fan beam collimator by applying a geometric transformation to each event
  • Parallel beam simulation is useful when a fan beam collimator is used to acquire an image of a small organ, but it is desirable to view an undistorted image which utilizes the image plane more efficiently
  • the image plane has a limited spatial resolution, to best utilize it, the region of interest should fill up as much of the image plane as possible.
  • an organ is imaged using a fan beam collimator
  • Each event is transformed by geometric transformer 78 to correct for the distortions caused by the fan beam collimator It should be noted that this transformation is applied before tomographic processing of the events, in the frame builder 78
  • variable mechanical misalignments are corrected in real-time.
  • detector 41 is moved along an axis of patient 40 If the speed of motion is not constant, portions of patient 40 which are scanned faster appear have less events than portions which are scanned more slowly In a preferred embodiment of the present invention, each event is given a weight depending on the actual speed of the scanning, a higher weight for fast scanning and a lower weight for slow scanning.
  • Fig 7 shows (in exaggerated form) another type of mechanical misalignment
  • Patient 40 has a Y axis along his body and an X-Z plane bisecting it
  • a nuclear camera 642 is rotated around patient 40, to capture events for forming a tomographic image
  • a first misalignment to be corrected is y-axis rotation misalignment Due to the great weight of camera 642, an actual center of rotation 602 is different from a configured center of rotation 600
  • camera 642 sags either towards or away from the center of rotation as a function of the rotation angle.
  • This misalignment is corrected in a preferred embodiment of the present invention by applying an angle dependent geometric transformation, to each event so that it is moved in the X-Z plane.
  • a second misalignment is axial (line of sight) miss-registration.
  • camera's 642 weight causes it to sag.
  • a nuclear medicine camera 644 is below him, it sags away from him.
  • the cameras are mounted on long beams, which bend so that the cameras viewing angle changes, the line of sight of the camera moves along the Y-axis of patient 40 as the amount of sagging changes.
  • this misalignment is corrected by applying an angle dependent geometric transformation to each event so that it is moved along the Y-axis.
  • Mechanical misalignments are typically detected either by configuration data, by feedback from mechanical position, velocity or acceleration sensors, or by processing the acquired data to find correlations and Fourier frequency peaks which correspond to certain types of misalignments.
  • y-axis rotation misalignment (described above), manifests itself as a Fourier frequency peak which corresponds to the degree of misalignment.
  • geometric transformations applied to an event are dependent on a measured biological rhythm, such as stomach contractions, breathing or heart beat.
  • processor 60 as described herein can be used with an analog front end, although a digital front end as described is preferred.

Landscapes

  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Optics & Photonics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Molecular Biology (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Nuclear Medicine (AREA)
  • Measurement Of Radiation (AREA)

Abstract

Ce procédé, qui sert à réduire les artefacts causés par la diffusion compton d'événements rayonnants dans une caméra à rayonnement (644), consiste à détecter un événement rayonnant, à déterminer son énergie et sa position, et à relier par convolution chaque événement sur un mode événement par événement avec une fonction de filtre spatial dépendant de l'énergie, qui est différente au moins en amplitude de la fonction de filtre à un second niveau d'énergie. L'opération de convolution est répétée pour plusieurs événements détectés.
PCT/IL1996/000162 1995-11-24 1996-11-24 Correction de la diffusion compton en temps reel WO1998023973A1 (fr)

Priority Applications (3)

Application Number Priority Date Filing Date Title
PCT/IL1996/000162 WO1998023973A1 (fr) 1996-11-24 1996-11-24 Correction de la diffusion compton en temps reel
JP52643398A JP2001516446A (ja) 1996-11-24 1996-11-24 リアルタイム・コンプトン散乱補正
US09/308,376 US6369389B1 (en) 1995-11-24 1996-11-24 Real-time compton scatter correction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/IL1996/000162 WO1998023973A1 (fr) 1996-11-24 1996-11-24 Correction de la diffusion compton en temps reel

Publications (1)

Publication Number Publication Date
WO1998023973A1 true WO1998023973A1 (fr) 1998-06-04

Family

ID=11061673

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IL1996/000162 WO1998023973A1 (fr) 1995-11-24 1996-11-24 Correction de la diffusion compton en temps reel

Country Status (2)

Country Link
JP (1) JP2001516446A (fr)
WO (1) WO1998023973A1 (fr)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000321357A (ja) * 1999-03-10 2000-11-24 Toshiba Corp 核医学診断装置
WO2004081606A1 (fr) * 2003-03-11 2004-09-23 Symetrica Limited Systeme de camera a rayons gamma ameliore
US10520613B2 (en) 2013-10-14 2019-12-31 Koninkluke Philips N.V. Histogram smoothing in positron emission tomography (PET) energy histograms
JP2020060545A (ja) * 2018-10-05 2020-04-16 キヤノンメディカルシステムズ株式会社 陽電子放出撮像装置及び方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105655435B (zh) * 2014-11-14 2018-08-07 苏州瑞派宁科技有限公司 光电转换器、探测器及扫描设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4424446A (en) * 1980-06-19 1984-01-03 Elscint, Ltd. Gamma camera correction system and method for using the same
EP0261696A2 (fr) * 1983-03-11 1988-03-30 Siemens Aktiengesellschaft Méthode de traitement d'impulsions utilisant la technique d'évaluation pondérée
US5438202A (en) * 1992-05-28 1995-08-01 Elscint Ltd. Stabilized scatter free gamma camera images
FR2741723A1 (fr) * 1995-11-24 1997-05-30 Elscint Ltd Systeme elabore de medecine nucleaire

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4424446A (en) * 1980-06-19 1984-01-03 Elscint, Ltd. Gamma camera correction system and method for using the same
US4424446B1 (en) * 1980-06-19 1994-04-19 Elscint Ltd Gamma camera correction system and method for using the same
EP0261696A2 (fr) * 1983-03-11 1988-03-30 Siemens Aktiengesellschaft Méthode de traitement d'impulsions utilisant la technique d'évaluation pondérée
US5438202A (en) * 1992-05-28 1995-08-01 Elscint Ltd. Stabilized scatter free gamma camera images
FR2741723A1 (fr) * 1995-11-24 1997-05-30 Elscint Ltd Systeme elabore de medecine nucleaire

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2000321357A (ja) * 1999-03-10 2000-11-24 Toshiba Corp 核医学診断装置
WO2004081606A1 (fr) * 2003-03-11 2004-09-23 Symetrica Limited Systeme de camera a rayons gamma ameliore
US7504635B2 (en) 2003-03-11 2009-03-17 Symetrica Limited Gamma-ray camera system
US10520613B2 (en) 2013-10-14 2019-12-31 Koninkluke Philips N.V. Histogram smoothing in positron emission tomography (PET) energy histograms
JP2020060545A (ja) * 2018-10-05 2020-04-16 キヤノンメディカルシステムズ株式会社 陽電子放出撮像装置及び方法

Also Published As

Publication number Publication date
JP2001516446A (ja) 2001-09-25

Similar Documents

Publication Publication Date Title
US6369389B1 (en) Real-time compton scatter correction
US5903008A (en) Scatter correction methods and systems in single photon emission computed tomography
Mosleh-Shirazi et al. A cone-beam megavoltage CT scanner for treatment verification in conformal radiotherapy
US5748768A (en) Method and apparatus for correcting distortion in an imaging system
US5128864A (en) Method for computing tomographic scans
US5376795A (en) Emission-transmission imaging system using single energy and dual energy transmission and radionuclide emission data
EP1006370B1 (fr) Méthode de reconstruction d'image en tomographie
US5293312A (en) Method and apparatus for computing tomographic scans
US5648997A (en) Apparatus and method for removing scatter from an x-ray image
US6381349B1 (en) Projector/backprojector with slice-to-slice blurring for efficient 3D scatter modeling
JP5172347B2 (ja) 反復的制約デコンボリューションによる核医学２次元平面画像の復元
US5818050A (en) Collimator-free photon tomography
Lewis et al. A megavoltage CT scanner for radiotherapy verification
Muehllehner et al. Design and performance of a new positron tomograph
JPH06258441A (ja) 放射線映像装置
JP2002511930A (ja) 二重モード・ガンマ・カメラ・システム
JPH0790024B2 (ja) Ｃｔ装置で画像を作成する方法
US10281596B2 (en) Correcting photon counts in a photon counting X-ray radiation detection system
JPH0838467A (ja) 検出器チャンネル利得較正係数を求める方法
US20230263499A1 (en) Counting response and beam hardening calibration method for a full size photon-counting ct system
US6921902B2 (en) Scatter correction device for radiative tomographic scanner
EP0195869B1 (fr) Filtre pour le traitement d'informations
JP2001099937A (ja) 医療撮像システムにおける放射汚染および不感時間損失の補正を行うための方法およびシステム
JPH10192267A (ja) Ｘ線装置
Karuta et al. Effect of detector weighting functions on the point spread function of high-resolution PET tomographs: a simulation study

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): JP US

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 09308376

Country of ref document: US

ENP Entry into the national phase

Ref country code: JP

Ref document number: 1998 526433

Kind code of ref document: A

Format of ref document f/p: F

122 Ep: pct application non-entry in european phase