CN101536913B - The method and apparatus of correcting multi-modality imaging data - Google Patents

Abstract

One is used for correcting the method (100) of PET (positron emission tomography) (PET) data and includes: tube current regulation (110) that will be generated by CT imaging system (10) to the second tube current value, and this second tube current value is less than the first tube current value for generating quality of diagnosis CT image；And utilization is set as that the CT imaging system of the second tube current value is to imaging patients.The method also includes: generate (102) multiple computerized axial tomography (CT) data for projection from CT imaging system；And pretreatment (104) CT data for projection is to generate pretreated CT data for projection.The method also includes: to pretreated CT data for projection filtering (106) to reduce electronic noise, thus generate filtered CT data for projection；And filtered CT data for projection is performed (108) negative logarithm operation to generate calibrated PET data.

Description

The method and apparatus of correcting multi-modality imaging data

Technical field

In general, the present invention relates to the imaging system that can operate in multiple modalities, more particularly, it relates to for the apparatus and method of correcting multi-modality imaging data.

Background technology

Multi-mode imaging system can utilize different mode to be scanned, and these mode are such as but not limited to PET (positron emission tomography) (PET), single photon emission computerized axial tomography (SPECT) and computerized axial tomography (CT).Conventional PET/CT imaging system obtains CT image to carry out correction for attenuation in PET and to provide anatomic information tracer used during helping doctor to be positioned at patient carries out PET imaging to take in.In some cases, CT diagnostic image is read respectively.The fully expression that the degree of accuracy of the CT value (CT number) rebuild is taken in for guaranteeing the tracer in organ is critically important.Under the normal condition of scanning, CT imaging system can easily meet this constraint, because the tolerance of the CT value coherence request needed for CT algorithm for reconstructing is the tightest.

In the multimodal systems of the most integrated PET/CT system, PET and CT image should record the most each other.When the detector of PET and CT imaging system is installed to public support body with physical method, intrinsic record can be produced.Conventional integrated PET/CT system or SPECT/CT system utilize the data generated by CT imaging system to come for PET or SPECT scan data and generate correction for attenuation information.Specifically, multiple transmitting correction for attenuation factors are derived from the CT data generated during CT scan.Term CTAC is for representing the transmitting attenuation quotient figure derived from CT image.The picture quality of CT diagnostic image is considerably beyond the picture quality generated needed for CTAC.

It addition, the imaging artefacts that patient moving causes becomes the most important problem for PET correction for attenuation.Such as, owing to CT image is typically to obtain in the short term, so the decay pattern that generated by CT image represents generally calls the attenuation characteristic of patient during gas in the motion that breathes no more.By contrast, PET image is typically to obtain within relatively long period, longer owing to obtaining the time, so allowing patient freely to breathe during this period.Not mating between both data acquisition, can cause occurring image artifacts in the PET image of correction for attenuation.

A kind of known method reducing imaging artefacts is that the CT image (or maximum intensity CT image) of multiple respiratory periods is average, to imitate the effect that the PET collected on multiple breathing cycles obtains.The another kind of known method reducing imaging artefacts is to utilize respiration gate control CT to obtain to generate the attenuation correction map that the respiratory characteristic obtained with respiration gate control PET more preferably mates.Typical agreement needs to utilize respiratory movement to monitor, and equipment obtains the 6-8 second to the CT that photographs.Then, the CT image of multiple respiratory period is rebuild so that the corresponding respiratory period in obtaining with respiration gate control PET mates.

Although every kind of known method all can efficiently reduce the image artifacts that patient moving causes, but owing to photography CT scan needs repeatedly to be continuously exposed to x-ray in certain prolonging period, so every kind of known method also can make the x-ray radiation dosage being delivered to patient increase.Because the resolution of diagnosis CT image is considerably beyond the resolution generated needed for CTAC, so a kind of method reducing x-ray radiation dosage is to reduce X-ray tube current, i.e. reduce CT image resolution ratio.

Such as, when X-ray tube current is set as the level that be enough to generate CT diagnostic image, the sub-fraction of the fluctuation that electronic noise is simply caused by x-ray photons statistics, and thus without the final projection reading of appreciable impact.But, when reducing X-ray tube current thus when reducing x-ray flux, it is a big chunk of overall noise that electronic noise becomes.Therefore, when reducing X-ray tube current to reduce patient dose, image shade can be made to dramatically increase, and CT value inaccuracy becomes readily apparent from.PET correction for attenuation unacceptable CT image is eventually produced for the level generating diagnosis CT image it addition, decreased below by X-ray tube current.

Summary of the invention

In one embodiment, it is provided that a kind of method of correcting multi-modality imaging data.The method includes: by the tube current generated by CT imaging system regulation to less than being used for generating the tube current value of the tube current value of quality of diagnosis CT image；And utilize the CT imaging system being set in relatively low tube current value to come imaging patients.The method also includes: generate multiple computerized axial tomography (CT) data for projection；And pretreatment CT data for projection is to generate pretreated CT data for projection.The method also includes: utilizes average to retain wave filter and to filter pretreated CT data for projection, to reduce the impact of electronic noise, and generates filtered CT data for projection；And filtered CT data for projection is performed negative logarithm operation (minus-logarithmic operation) to generate calibrated multi-modality imaging data.The method also includes generating CT image, uses this CT image to generate the correction for attenuation factor for correcting PET data.

In another embodiment, it is provided that a kind of multi-mode imaging system, this system includes first mode unit, second mode unit and is operatively coupled to the computer of the first and second modality units.This computer programming is for generating multiple computerized axial tomographies (CT) data for projection, and these CT data for projection of pretreatment are to generate pretreated CT data for projection.This computer is also programmed to the filtering of pretreated CT data for projection to reduce the impact of electronic noise, and filtered CT data for projection performs negative logarithm operation and other calibration steps to generate calibrated multi-modality imaging data.

In yet another embodiment, it is provided that a kind of with being programmed for indicating the computer-readable medium of the program coding of computer.This computer-readable medium is programmed for generating multiple computerized axial tomography (CT) data for projection, and these CT data for projection of pretreatment are to generate pretreated CT data for projection.This computer-readable medium is also programmed to reduce electronic noise, the filtering of pretreated CT data for projection is generated filtered CT data for projection, and filtered CT data for projection performs negative logarithm operation and other calibration steps to generate calibrated multi-modality imaging data.

Accompanying drawing explanation

Fig. 1 is the diagram of the most exemplary multi-mode imaging system.

Fig. 2 is the schematic block diagram according to the system shown in one embodiment of the invention, Fig. 1.

Fig. 3 A and 3B is the flow chart illustrating the exemplary method for correcting PET (positron emission tomography) (PET) data.

Fig. 4 is that the figure according to one embodiment of the invention, corresponding to one group of CT projection value of the single detector row of single projection view represents.

Detailed description of the invention

Fig. 1 is the diagram of the most exemplary multi-mode imaging system 10.Fig. 2 is the schematic block diagram according to the multi-mode imaging system 10 shown in one embodiment of the invention, Fig. 1.Although include CT imaging system and PET imaging system exemplary bimodal imaging system context described in embodiments of the invention, it will be appreciated that, it is contemplated that other imaging system being able to carry out functions described herein can be used.

Provided herein is a kind of system and method for correcting multi-modality imaging data.These apparatus and method are described with reference to the accompanying drawings, and in all figures, similar numeral represents identical element.These figures are used for illustrating and noting limit, and are included in this so that explanation.

As it has been described above, multimodal systems known at least one utilizes CT diagnostic quality images, this had both been in order at the purpose of diagnosis, and for generate the multiple correction for attenuation factors derived from CT data.In such multimodal systems of the most integrated PET/CT system, there is the intrinsic record of PET and the CT image that system obtains.Lie on same workbench because patient is quiet during PET and the CT part obtained, so during this twice acquisition, patient will be located in uniform position and orientation, which greatly simplifies process that is interrelated to CT and PET image and that merge.This allows to use CT image to come for the reconstruction of PET image and provides correction for attenuation information, and allows image solution reader easily by interrelated for the function information that presents in the anatomic information presented in CT image and PET image.It is desirable, however, that utilize ratio for generating the little X-ray tube current of the X-ray tube current of CT diagnostic quality images or utilizing ratio to provide dampening information for the reconstruction of PET image for generating the low x-ray tube voltage of the x-ray tube voltage of diagnostic quality images.Also want to reduce and/or eliminate the imaging artefacts occurred in CT image when reducing x-ray tube current.

Seeing figures.1.and.2, it is shown that multi-mode imaging system 10, this system 10 includes first mode unit 11 and second mode unit 12.The two modality unit 11 and 12 makes system 10 can utilize first mode unit 11 sweep object or patient 22 at first mode, and can utilize second mode unit 12 sweep object at second mode.System 10 allows to take multiple scan in different mode, is beneficial to make diagnosis capability increase than single mode system.In one embodiment, multi-mode imaging system 10 is PET (positron emission tomography)/computerized axial tomography (PET/CT) imaging system 10.Alternatively, system 10 uses the mode beyond CT and PET.First mode unit 11 (that is, CT imaging system) includes the frame 13 with x-radiographic source 15, and this x-radiographic source 15 detector array 18 on the opposite being positioned at frame 13 projects x-beam 16.Detector array 18 is formed by multiple detector row, and each detector row includes the multiple detector elements 20 jointly sensing the projection x-ray of the object through such as medical patient 22.Each detector element 20 produces the signal of telecommunication of the intensity representing incident X-ray bundle, and therefore allows to estimate the decay of X-ray bundle when X-ray bundle is through object or patient 22.During being scanned obtaining x-ray projection data, frame (gantry) 13 and assembly attached thereto rotate around center of rotation 24.Fig. 2 only illustrates single-row detector element 20 (that is, one detector row).But, many multi-slice detector array 18 includes the multiple parallel detector row being made up of detector element 20, so that the data for projection corresponding to multiple sections can be obtained during scanning simultaneously.

The operation with x-radiographic source 15 that rotates of frame 13 is controlled by the controlling organization 26 of PET/CT system 10.Controlling organization 26 includes for providing the x-ray controller 28 of power and timing signal and the gantry motor controller 30 of rotary speed and position for controlling frame 13 to x-radiographic source 15.Analog data from detector element 20 is sampled by the data-acquisition system (DAS) 32 in controlling organization 26, and converts this data to digital signal for post processing.Image reconstructor 34 receives sampling and digitized x-ray data from DAS 32, and performs high speed image reconstruction.With the image rebuild as the input of computer 36, computer 36 stores the image in storage device 38.Computer 36 also receives order and sweep parameter via the control station 40 with keyboard from operator.The visual display unit 42 of association allows operator to observe the reconstruction image from computer and other data.

Computer 36 utilizes order that operator provide and parameter to provide control signal and information to DAS 32, x-ray controller 28 and gantry motor controller 30.Additionally, computer 36 also operative employee's station motor controller 44, this controller 44 controls electric table 46 patient 22 to be positioned in frame 13.Specifically, patient 22 is moved through frame openings 48 by workbench 46 at least partially.

In one embodiment, computer 36 includes equipment 50, such as floppy disk, CD-ROM drive, DVD drive, magneto-optic disk (MOD) equipment or other digital device any including the network access device of the such as ethernet device for reading instruction and/or data from computer-readable medium 52, other digital source of computer-readable medium 52 such as floppy disk, CD-ROM, DVD or such as network or the Internet and the digital unit that will develop.In another embodiment, computer 36 performs to be stored in the instruction in firmware (not shown).Computer 36 is programmed to carry out functions described herein, as used herein, term " computer " is not limited solely to those integrated circuits being referred to as computer in the art, but broadly referring to computer, processor, microcontroller, pico computer, programmable logic controller (PLC), special IC and other programmable circuit, these terms are used interchangeably herein.

PET/CT system 10 also includes the second image mode 12, i.e. PET image mode.PET image mode 12 includes PET emitter-detector 60, and this detector 60 includes multiple detector element (not shown).PET emitter-detector 60 and CT detector array 18 all detects radiation, and at referred herein to radiation detector.In one embodiment, PET/CT system 10 be can from Waukesha Wis., the Discovery LS PET/CT system buied of General Electric Healthcare, and configure as described.In another embodiment, multi-mode imaging system 10 be equally can from Waukesha Wis., the Hawkeye PET/SPECT/CT system buied of GE Healthcare, and configure as described.It addition, be described despite in medical context, it is contemplated that invention described herein also can make all multi-mode imaging systems be benefited.

In an exemplary embodiment, x-radiographic source 15 and detector array 18 rotate around patient 24 to be imaged together with frame 13 in imaging plane, so that the angle that x-beam 16 intersects with patient 24 constantly changes.A frame angle from detector array 18 one group of x-ray attenuation measured value (that is, data for projection) be referred to as one " view (view) "." scanning " of patient 24 is included in x-radiographic source 15 and detector 18 once rotate during one group of view being taken at different frame angles or visual angle.

In CT scan, process data for projection to build corresponding to the image to the two dimension slicing that patient 24 is taken.A kind of method from one group of data for projection reconstruction image is referred to as filtered back projection technique in the art.Overall attenuation measured value is converted to represent the image of the decay of patient in each pixel by this process.Generally attenuation measurements is converted to CT value unit or Huo Sifeierde unit (Hounsfield unit).

In order to reduce total sweep time, " spiral " scanning can be performed.In order to perform " spiral " scanning, mobile patient 24, obtain the data of the section of specified quantity simultaneously.This system generates single spiral from fan-shaped beam helical scanning.The spiral drawn by fan-shaped beam produces data for projection, from these data for projection, can rebuild the image in each prescribed slice.Many section (multi-slice) detectors are utilized to obtain multiple spirals.

The algorithm for reconstructing of helical scanning generally utilizes helical weighing algorithms, the function that this algorithm indexes according to visual angle and the probe access data weighting to being gathered.Specifically, before filtered back projection's process, according to the helical weighing factor weight data of the function as frame angle and probe angle.Then, weighted data is processed to generate CT value and to build corresponding to the image to the two dimension slicing that patient 24 is taken.In the operating process of many section PET/CT systems 10, utilize multiple detector row to obtain multiple projection simultaneously.Similar with the situation of helical scanning, before filtered back projection's process, to data for projection application weighting function.

In the operating process of PET imaging system 19, the radionuclide emission utilizing cyclotron or miscellaneous equipment to prepare goes out positron, i.e. the electronics of positively charged or positron.The radioactive nucleus being most frequently with in diagnosing image have Value linear (18F), rubidium-82 (82Rb), carbon-11 (11C), nitrogen-13 (13N) and oxygen-15 (15O).The material of such as glucose or carbon dioxide use radionuclide referred to as the radioactive indicator of " radiopharmaceutical " by being attached to by radionuclide.

In generally using, by radiopharmaceutical injection to the patient of such as patient 24, this medicine will built up in the organ of imaging, blood vessel etc..It is known that specific radiopharmaceutical is concentrated in some organ, or in the case of blood vessel, this radiopharmaceutical will not be absorbed by blood vessel wall.Centralized procedure generally includes such as glucose metabolism, fatty acid metabolism and the process of protein synthesis.

After radiopharmaceutical concentrates in organ of interest, while radionuclide is changed in quality, radionuclide emission positron.Positron just runs into electronics after the distance that experience is the shortest, and when positron runs into electronics, positron can fall into oblivion and be converted to two photons.This annihilation events can be characterized by two features that the imaging of medical with imaging, especially with PET (positron emission tomography) (PET) is relevant.First, when falling into oblivion, the energy of each annihilation photon is about 511keV.Second, the two annihilation photon is along essentially the inverse direction orientation.

In PET imaging, if the Position Approximate can fallen into oblivion three dimension identifications, then can rebuild the 3-D view of organ of interest for observing.In order to detect annihilation position, use PET camera, such as emitter-detector 60 (as shown in Figure 2).Exemplary PET camera includes multiple detector and a processor, also includes (coincidence) detection circuit that overlaps.

Coincidence circuit (not shown) identification pulse pair substantially simultaneously, these pulses are to the detector corresponding to being positioned essentially on the opposite side of imaging region.Therefore, pulse simultaneously, to instruction, the straight line between the detector of a pair association there occurs annihilation events.Within acquisition period of a few minutes, record millions of times and fall into oblivion, and will fall into oblivion with specific detector being associated via such as PET DAS 32 every time.After acquisition period, can via several different known to back projection during any one method utilize the annihilation data recorded to build the 3-D view of organ of interest.

As used herein, unless clearly described, otherwise should and will be interpreted as being not excluded for a plurality of described element or step titled with element or the step of prefix by word "a" or "an" with odd number narration.Additionally, be not to be construed as getting rid of the existence of the Additional examples of composition also including described feature when mentioning " embodiment " of the present invention.

The most as used herein, phrase " reconstruction image " is not meant to get rid of wherein generate and represents that the data of image do not generate the embodiments of the invention of visual image.Therefore, as used herein, term " image " broadly refers to visual image and represents the data of visual image.But, many embodiments generate (or being disposed for generating) at least one visual image.

As described previously herein, the data that PET/CT system 10 is configured with by the CT mode of operating system 10 generates generate the correction for attenuation factor, can utilize these correction for attenuation factors during the PET mode of operating system 10.Specifically, derive from the CT data generated during CT scan and launch the correction for attenuation factor (CTAC), wherein CT system is configured to produce the spacing of optimal CT picture quality and technology generates CT data, and utilizes identical CT data for projection to generate the PET image through correction for attenuation.

During operation PET/CT imaging system 10, patient 24 is scanned generating more than 102 CT image.As discussed herein, term " image " broadly refers to visual image and represents the data of visual image.Then, the image generated is utilized to generate the CT correction for attenuation factor.

Fig. 3 is the flow chart illustrating the exemplary method 100 for correcting multi-modality imaging data.In an exemplary embodiment, multi-modality imaging data can be such as PET or SPECT imaging data.In an exemplary embodiment, method 100 includes: generate more than 102 computerized axial tomography (CT) data for projection；And pretreatment 104CT data for projection is to generate pretreated CT data for projection.Method 100 also includes: to the filtering 106 of pretreated CT data for projection to reduce electronic noise, thus generate filtered CT data for projection；And filtered CT data for projection is performed 108 negative logarithm operations to generate calibrated PET data.

In order to generate CT diagnostic quality images, method 100 includes tube current regulation 110 to the second tube currents used by CT imaging system to generate attenuation correction map, and wherein the second tube current is less than the first tube current.More specifically, CT imaging system 11 is configured to when tube current level is set as the first tube current generate CT quality diagnosis image, and in an exemplary embodiment, the first tube current is between about 200mA-400mA.But, as it has been described above, the resolution of CT diagnostic quality images is considerably beyond the resolution created needed for CTAC.Therefore, method 100 includes the level adjustment 110 of tube current or to reduce 110 be herein referred as attenuation correction map (CTAC) value or the value of the second tube current value, and to generate attenuation correction map, wherein the second tube current is less than the first tube current value.

CTAC value is less than the tube current setting value for generating quality of diagnosis CT image.Such as, the second tube current value or attenuation correction map (CTAC) value are less than 20mA.In an exemplary embodiment, attenuation correction map (CTAC) value is between about 10mA and about 20mA.Tube current being set as, attenuation correction map (CTAC) value is beneficial to reduce total x-roentgendosis that patient 22 received between sweep time, is additionally favorable for reducing CT image resolution ratio, this is discussed below.Method 100 is additionally included in the case of tube current is set as attenuation correction map (CTAC) value and generates more than 102 computerized axial tomography (CT) data for projection.

Method 100 also includes that pretreatment 104CT data for projection is to generate pretreated CT data for projection.In an exemplary embodiment, pretreatment 104CT data for projection includes CT data for projection is applied 112 detector gain calibrations, reference channel normalization and/or other correction.Alternatively, pretreatment 104CT data for projection may also include application scatter correction or utilizes other technology various to carry out pretreatment CT data for projection.

As it has been described above, reduction tube current value or magnitude of voltage can cause negative signal.Such as, the figure of one group of CT projection value of single detector row of the single projection view that Fig. 4 corresponds to anthropometric dummy represents.As shown in Figure 4, when x-ray tube current is set as attenuation correction map (CTAC) value, a large amount of passages drop to less than zero.These negative values cause a large amount of shade and striped occur in the image rebuild.

These negative values are because there is electronic noise in CT DAS system 32 and causing.As discussed above, when tube current is set as generating diagnostic quality images, i.e. when tube current is set as of a relatively high tube current, of a relatively high tube current causes of a relatively high x-ray flux situation occur.Because the sub-fraction of the fluctuation that electronic noise is simply caused by x-ray photons statistics, so electronic noise is not significantly affected by final projection reading.But, when tube current is set as the value lower than the value being used for producing diagnosis or Anatomical orientation acceptable CT image, x-ray flux can reduce, and electronic noise to become be a big chunk in overall noise, thus result in negative value as shown in Figure 4.In order to reduce the impact of electronic noise, method 100 also includes pretreated CT data for projection filtering 106 to reduce the impact of electronic noise, thus generates filtered CT data for projection.

The filtering carried out to remove the negative value producing striped in diagnostic quality images can by with near probe access pre-processed projection data meansigma methods replace negative value realize.Replacing negative value can make average or meansigma methods change and become more positive value, and the data obtained would indicate that less decay.For the scanning performed at relatively low tube current, this change can make CT value offset to less value, and the CTAC value of generation can be made PET data undercorrection.In order to generate accurate CTAC data, retain average for the filtering or other process removing negative value.

In an exemplary embodiment, pretreated CT data for projection filtering 106 is included to reduce electronic noise: utilize low pass filter that pretreated CT data for projection filters 106, then CT data for projection is performed negative logarithm operation.Following formula represents the exemplary low pass filter that can be used for filtering pretreated data for projection.

$p^{\′}(\mathrm{\γ},\mathrm{\β},z)=-\log (p(\mathrm{\γ},\mathrm{\β},z)\⊗f(\mathrm{\γ},\mathrm{\β},z))$ Formula 1

Wherein γ, β and z represent probe access, view and line index respectively, f (γ, β, z) be low pass filter, andIt it is convolution operator.

In an exemplary embodiment, before negative logarithm operation and after carrying out detector gain calibration, filter step 106 can be performed at many diverse locations.Specifically, in the case of not carrying out detector gain calibration, projection signal exists probe access-channel gain change, and low-pass filtering operation can mix these changes in filtering signal, thus in final image, produce ring artifact.Once application suitably corrects the inhomogeneities in each dimension, just can perform the smoothing processing of this dimension.View, view smoothing processing can perform before or after detector gain is calibrated, and passage-passage smoothing processing should perform after air calibration.Each dimension independently executes smoothing processing can easily accessible suitable data.

In use, utilize low pass filter that pretreated data for projection filtering 106 is reduced spatial resolution.But, method 100 utilizes not mating between CT image and the spatial resolution of PET image.As it has been described above, the spatial resolution of CT is far above PET, and considerably beyond the necessary resolution of decay pattern.Therefore, pretreated CT data for projection is performed 106 filtering operations and do not interfere with the degree of accuracy of final decay pattern.More precisely, utilize restitution nucleus (kernel) and post-processing technology specially to reduce the spatial resolution of CT image.

In one embodiment, also include utilizing three-dimensional rectangle wave train (boxcar) wave filter that CT data for projection is filtered 120 to reduce electronic noise to CT data for projection filtering 106.Utilize low-pass filtering operation, boxcar smoothing processing, can formula (1) is reduced to as follows:

$p^{\′}(\mathrm{\γ},\mathrm{\β},z)=-\log \left(\frac{1}{(2N+1)(2M+1)(2L+1)}\underset{\mathrm{\Δ\γ}=N}{\overset{N}{\mathrm{\Σ}}}\underset{\mathrm{\Δ\β}=M}{\overset{M}{\mathrm{\Σ}}}\underset{\mathrm{\Δz}=L}{\overset{L}{\mathrm{\Σ}}}p(\mathrm{\γ}+\mathrm{\Δ\γ},\mathrm{\β}+\mathrm{\Δ\β},z+\mathrm{\Δz})\right)$ Formula 2

Wherein γ, β and z represent probe access, view and line index respectively, and (γ, β are z) to perform the average or three-dimensional low pass boxcar filter of filtering at view, row and channel dimension, and N, M and L are boxcar filter parameters p '.These parameters can predefine based on scanning technique and sweep object, or can dynamically determine based on the amount of negative value present in projection.

Alternatively, include utilizing to CT data for projection filtering 106 and independently execute that 5 × 5 × 5 or 5 × 3 × 5 or 3 × 1 × 3 is average or the wider core of the three-dimensional low pass boxcar filter of filtering comes projection filtering 120 the most respectively at view, row and channel dimension.Such as, exemplary wave filter can be 3 × 3 wave filter in 3 × 1 wave filter on view direction, 5 × 1 wave filter on line direction and channel direction.In an exemplary embodiment, include utilizing the low pass filter of any size, the core coefficient such as with weighted mean and/or other desired low pass filter any to CT data for projection filtering 120, and can have different core in each dimension.In an exemplary embodiment, if still observing negative value in gained signal, then the average that filtered market demand is other is retained filtering.

In another embodiment, include utilizing different data acquisition available in CT scanner that CT data for projection is filtered 122 to CT data for projection filtering 106, in order to perform the equivalent operation as performed by method 120.In an exemplary embodiment, filtering 122 includes: obtain CT data for projection in thicker detector row；And by average for the electronic noise in multiple input channels.Such as, 16 × 1.25mm detector can be configured to 16 × 1.25mm detector, 8 × 2.5mm detector or 4 × 3.75mm detector.In this embodiment, filtering 122 includes utilizing 8 × 2.5 or 4 × 3.75mm detector configurations to obtain CT data for projection.Because in FEP, analog prober signal can be sued for peace before pre-treatment step 104 and therefore before performing 108 negative logarithm operations, so after data acquisition completes, it is not necessary to independent process just can realize by row filtering.In an exemplary embodiment, the analogue signal utilizing 4 × 3.75 detectors configurations to generate is about utilizing three times of the analogue signal that 16 × 1.25 detectors configurations generate big, and this Approximate Equivalent utilizes the core of a length of 3 to be averaged in dimension of being expert at as mentioned above.Additionally, utilize the configuration of 16 × 1.25mm detector to utilize whole 20mm of detector array 18 when obtaining data, and the configuration of 4 × 3.75mm detector is utilized when obtaining data, only to utilize the 15mm of detector array 18.Alternatively, CT data for projection may utilize the configuration acquisition of 8 × 2.5mm detector.Whole 20mm of 8 × 2.5mm detector configuration using detector array 18, this Approximate Equivalent utilizes the core of a length of 2 to be averaged in dimension of being expert at as mentioned above, thus the analogue signal that the twice of the generation about analogue signal of 16 × 1.25mm detector array configuration is big.

The detector configuration utilizing the 16 × 1.25mm detector configuration being different from the whole 20mm utilizing detector array 18 provides the method more effectively filtered CT data for projection.Such as, the projection sample from different detector row is processed by different processors, almost without communication path between these processors.Therefore, detector row, probe access or projection view are performed average or low-pass filtering is more time-consuming.But, utilize the configuration of 4 × 3.75mm detector or as above the configuration of other optional detector allow imaging system to utilize the data obtained in identical view and row to process CT data for projection, thus add and probe access performed the average or efficiency of low-pass filtering.In another embodiment, probe access has the projection of thicker spatial resolution across the complete arrangement of detector arc (gang) with generation.Such as, two adjacency channels electronically combine, in order to realize 888 passage × 64 row of 444 passage × 64 row data acquisitions rather than nominal.The method also can be with detector row configuration combination to make each view produce 444 passage × 16 row data samples.Certainly, other detector configuration combination is also possible.

In another embodiment, CT data for projection filtering 106 is included, and utilizing view compress technique to come CT data for projection filters 124.In an exemplary embodiment, view compression is such a pattern, wherein frame with than normal acquisition time slow-footed speed rotate so that rotating the view producing integral multiple when quantity is normally to obtain every time.During pre-treatment step 104, before the negative logarithm of application, these views are average, to reduce the noise in rebuilding.In an exemplary embodiment, in order to realize the low-dose CT of PET correction for attenuation, utilize identical gantry speed to obtain CT projection view (that is, amount of views is identical with amount of views during for generating CT diagnostic quality images).It addition, before the negative logarithm of application, utilize view compress technique to reduce amount of views, then by average for the electronic noise component of signal.In use, view compress technique is to CT data for projection application view compressibility factor.In an exemplary embodiment, view compressibility factor is 3: 1, it is achieved thereby that perform the effect of the equivalent of filtering with the verification view utilizing as discussed above a length of 3.Therefore, method 100 is beneficial to generate accurate CT Huo Sifeierde unit value when utilizing low dosage x-ray scanning patient.In another embodiment of the present invention, DAS (data-acquisition system) sampling rate changes, so that it is slower than diagnosis CT data acquisition sampling rate.Such as, for the gantry rotation rate of a second, DAS sampling rate is decreased to 500Hz rather than 1000Hz, so that the quantity of the x-ray photons of each sample increases.

After being filtered reducing electronic noise to pretreated CT data for projection, method 100 also includes filtered CT data for projection is performed 108 negative logarithm operations.In an exemplary embodiment, negative logarithm step includes that the extra process such as taking the absolute value of projection value is to eliminate any negative detector signal.Or, average can be applied to retain wave filter, in order to the average signal after making logarithm operation is basically unchanged.Average retains wave filter will give one predetermined value of negative sample value, change the value of its adjacent sample to make average substantially keep identical simultaneously.Or, negative value can be replaced by the value of dynamic calculation rather than predetermined value, and correspondingly revise adjacent channel value.Then, treated data for projection is fed to other pretreatment and reconstruction procedures and rebuilds image to obtain.

Method described herein and device are easy to generate the correction for attenuation data that can be used for removing the impact of decay and scattering from PET image.Specifically, method described herein and device are easy to reduce the image artifacts that patient moving causes, and are also convenient for reducing the x-ray dose being delivered to patient simultaneously.In order to overcome the shade and other imaging problem typically associated with low dose radiation imaging, the method filters (filter) to CT data for projection before being additionally included in the negative logarithm operation of execution, the negative projection value caused due to the electronic noise in CT data-acquisition system with minimizing.

Although describing the present invention with regard to each specific embodiment above, but those skilled in the art is it will be appreciated that in the spirit and scope of the claims, it is achieved can make amendment during the present invention.

The above-detailed one exemplary embodiment of multi-modal (multi-modality) imaging system.The assembly of shown multi-mode imaging system is not limited to specific embodiments described herein, but can the most independently utilize the assembly of each multi-mode imaging system with other assembly described herein.Such as, may also be combined with other imaging system and use above-described multi-mode imaging system assembly.

As used herein, term " computer " can include any based on processor or system based on microprocessor, including utilizing microcontroller, Reduced Instruction Set Computer (RISC), special IC (ASIC), logic circuit and other circuit being able to carry out functions described herein any or the system of processor.Above example is exemplary, in any case be therefore not meant to limit definition and/or the implication of term " computer ".

Computer or processor perform the one group of instruction being stored in one or more memory element, in order to process input data.Memory element is also dependent on being desired or needed for storing data or out of Memory.Memory element can be to be the information source in handling machine or the form of physical memory element.

The various orders of the specific operation of this method organizing each embodiment that instruction can include indicating the computer as handling machine or processor to perform the such as present invention and process.This group instruction can be to use the form of software program.Software can use various forms, such as systems soft ware or application software.Additionally, software can also use the form of a part for the program module in the set of stand-alone program, large program or program module.Software may also include the module programming using OOP form.Handling machine can respond user command or responds previously processed result or respond request that another handling machine done and carry out the process of input data.

As used herein, term " software " and " firmware " can exchange, and including storing any computer program performed in memory for computer, memorizer includes RAM memory, ROM memory, eprom memory, eeprom memory and non-volatile ram (NVRAM) memorizer.Above memorizer is similar the most exemplary, and therefore unrestricted about the type of memory that can be used for storing computer program.

It will be appreciated that above description is only intended to illustrate and not limit.Such as, above-described embodiment (and/or its aspect) can be in combination with one another.It addition, in the case of without departing substantially from the scope of the present invention, many amendments can be made to make particular condition or material be suitable for the teachings of the present invention.While characterized as the size of material and the parameter of the type present invention to be defined, but they are the most restrictive, and simply one exemplary embodiment.After reading the above description, it will be appreciated by those skilled in the art that many other embodiments.Therefore, it should with reference to the four corner of appended claims and the equivalent giving these claim to determine the scope of the present invention.In appended claims, term " includes (including) " and " wherein (in which) " is used as corresponding term and " comprises " and the popular equivalence word of " wherein (wherein) ".Additionally, in appended claims, term " first ", " second " and " the 3rd " etc. are used merely as labelling rather than their object are applied numerical requirements.Additionally, means-plus-function format writing is not used in the restriction of appended claims, use phrase " to be used for ... parts (means for) " adding function statement and there is no further structure clearly unless these claim limits, be otherwise not meant to according to 35U.S.C. § 112 the 6th section, it be explained.

Claims (6)

1. for an equipment for correcting multi-modality imaging data, including:

For the tube current used by computerized axial tomography CT imaging system is regulated (110) to first Tube current value is to generate the device of attenuation correction map, and described first tube current value is not enough to generate diagnosis Quality CT image；

Many for generating (102) from the described CT imaging system being set in described first tube current value The device of individual CT data for projection, wherein said first tube current value causes described CT data for projection In negative value；

Number is projected to generate pretreated CT for pretreatment (104) described CT data for projection According to device；

For to described pretreated CT data for projection filtering (106) to reduce electronic noise, Thus generate the device of filtered CT data for projection；And

For described filtered CT data for projection being performed (108) negative logarithm operation to generate warp The device of the multi-modality imaging data of correction, described negative value is converted to by wherein said negative logarithm operation On the occasion of,

The wherein said equipment for correcting multi-modality imaging data farther includes: be used for utilizing low Logical boxcar filter carrys out the device to the filtering of described pretreated CT data for projection, or uses Filter in utilizing three-dimensional low pass boxcar filter to come described pretreated CT data for projection Device, or for described pretreated CT data for projection apply 3: 1 view compressibility factors Device.

2. equipment as claimed in claim 1, also includes for described filtered CT projection Data perform negative logarithm operation to generate calibrated PET data and calibrated SPECT data At least one of device.

3. equipment as claimed in claim 1, also includes for only at passage, row and view direction On the spot application averaging operator wave filter is with the dress filtering described pretreated CT data for projection Put.

4. equipment as claimed in claim 1, also includes for utilizing 4 × 3.75 millimeters of detectors Configuration obtains the device of described CT data for projection.

5. the method (100) for correcting multi-modality imaging data, described method includes:

Tube current used by computerized axial tomography CT imaging system is regulated (110) to the first pipe electricity Flow valuve is to generate attenuation correction map, and described first tube current value is not enough to generate quality of diagnosis CT figure Picture；

At described first tube current value, generate (102) multiple CT from described CT imaging system throw Shadow data, wherein said first tube current value causes the negative value in described CT data for projection；

Pretreatment (104) described CT data for projection is to generate pretreated CT data for projection；

To described pretreated CT data for projection filtering (106) to reduce electronic noise, thus Generate filtered CT data for projection；

Utilize low pass boxcar filter that described pretreated CT data for projection is filtered Ripple, or utilize three-dimensional low pass boxcar filter to come described pretreated CT data for projection It is filtered, or described pretreated CT data for projection is applied 3: 1 view compressibility factors； And

Described filtered CT data for projection is performed (108) negative logarithm operation calibrated to generate Multi-modality imaging data, wherein said negative logarithm operation described negative value is converted on the occasion of.

6. method (100) as claimed in claim 5, also includes described filtered CT Data for projection performs (108) negative logarithm operation to generate calibrated PET data and calibrated At least one of SPECT data.