Patent Application
20090242776
The present invention relates to a preparation method for image information by an image processing device.
A PET device has been used, as an important tool, mainly in an object of diagnosis of malignant tumor in a medical field, in recent years. The PET device measures radiation (γ-ray) emitted from the inside of a patient body, derived from a radiopharmaceutical injected into a patient as a subject body, and images distribution of the radiopharmaceutical in the patient body from measurement data thereof. Such a PET device is used in diagnosis of metabolic function and physiologic function. As a typical radiation image diagnosis device, which applies radiation measurement technology, for non-invasively obtaining image of the inside of the patient, there is an X-ray CT device.
The PET device is equipped with many radiation detectors, and is required to process a huge amount of radiation detection signals (γ-ray detection signals), which are output from these radiation detectors. Quality of obtained image information has been limited from restriction of data processing, because image information (PET image information) is reconstructed by processing of the huge amount of data obtained by these radiation detection signals. However, with rapid progress of a signal processing circuit and computer technology, in recent years, image information with high quality has enabled to be provided.
Diagnosis of malignant tumor using the PET device is executed as follows. Firstly, the radiopharmaceutical, labeled with a positron emission nuclei species (15O, 13N, 11C, 18F and the like), which specifically accumulates at a specific moiety in a body (hereafter called a PET pharmaceutical), is administered to the patient. A positron, which is emitted from the PET pharmaceutical in the patient body, experiences positron annihilation by binding with an electron of inside of a nearby cell. In this annihilation, a pair of γ-rays (hereafter called pair γ-rays) having an energy of 511 KeV are emitted. Because each of the pair γ-rays is emitted in a nearly opposite direction each other, it can be specified by coincidence counting of both γ-rays, at what position inside the body the positron annihilation event occurred. These γ-rays are detected with the radiation detectors. After the detection of statistically sufficient number of the pair γ-rays, occurrence frequency distribution of the pair γ-rays, that is distribution of the PET pharmaceutical inside the patient body, can be imaged, by using image information reconstruction algorism, such as a filtered-back-projection method (refer to IEEE Transactions on Nuclear Science, Vol. NS-21, pages 228 to 229). Measurement of the γ-rays to be generated, caused by the PET pharmaceutical inside the body, is called emission measurement (hereafter called E measurement), and image information reconstructed, based on γ-ray detection signals obtained by the emission measurement, is called emission image information (hereafter called E image information). The E image information is generally called simply PET image, however, in the present description, to differentiate from transmission image information (hereafter called T image information) to be described later, it is called E image information. In addition, a series of processes from E measurement to reconstruction is called emission imaging collectively.
In an inspection using such a PET device, for example, in the case where the patient is administered with a PET pharmaceutical called FDG (Fluoro-2-deoxyglucose), which is an analog of a saccharide (glucose), the PET pharmaceutical is accumulated at malignant tumor having larger saccharide metabolism as compared with a normal moiety. Therefore, diagnosis of position and shape of malignant tumor becomes possible.
Incidentally, in an inspection using the PET device (PET inspection), which requires quantitativeness, in addition to E measurement, there is also executed measurement called transmission (the transmission measurement, hereafter called T measurement), by using a γ-ray source, which is a transmission radiation source installed at the PET device. Attenuation of γ-ray in the PET measurement indicates phenomenon where γ-ray derived from the radiopharmaceutical is not detected as coincidence count data effective in imaging, as a result of impact of interaction thereof with substances inside a body, till emission outside a patient body. A process to compensate this attenuated amount of this γ-ray is called attenuation compensation, and at present, it is executed in most of the PET inspections.
The attenuation compensation usually uses data obtained by T measurement. That is, by rotating a γ-ray source around the patient lying on a bed, γ-ray (radiation) emitted from the γ-ray source is irradiated onto the patient. Each of radiation transmittances in various directions, which this γ-ray transmits through the patient, is determined. Data obtained by E measurement is compensated by using data of these radiation transmittances. As the γ-ray source, a radio-isotope (hereafter called “RI”) such as 68Ge—68Ga and 137Cs is usually used. Instead of the γ-ray source, an X-ray source, described in JP-A-2006-231083, may also be used. It should be noted that, if necessary, tomographic image information of the patient is reconstructed, based on data obtained by T measurement. This tomographic image information is morphological image, and is called hereafter T image information. The T image information is one representing distribution of radiation attenuation inside a patient body. If necessary, attenuation rates by each projection direction may be determined again based on the image information, and be used for attenuation compensation of these attenuation rates.
In recent years, a combined PET/CT device has been proposed, where an X-ray CT device is arranged in parallel to and combined with a PET device. This combined PET/CT device executes attenuation compensation by utilization of tomographic image information obtained by the X-ray CT device. JP-A-2006-231083 has proposed a PET device with a structure rotating an X-ray source inside a plurality of radiation detectors arranged circularly. This PET device also executes attenuation compensation by utilization of tomographic image information reconstructed by using detection signals of X-rays emitted from the X-ray source and transmitted through the patient.
As a large factor to decrease quality of E image information, impact of movement of a patient (hereafter called patient motion) is included. In the patient motion, there is periodic movement accompanied with involuntary respiration and heart beat, and a voluntary posture change. Because the PET inspection requires a long measurement time, usually from several minutes to several-ten minutes, it is difficult to suppress patient motion, without giving stress to the patient. In particular, because motion by respiration (hereafter called respiration motion) reaches as large as 2 to 3 cm, even in resting respiration, impact of patient motion giving on E image information is large in the PET inspection of a moiety near a lung field.
As a method for compensating blurring of image information accompanied with periodical patient motion such as respiration, a method called gated-acquisition has been known. The gated-acquisition is a method for dividing data of E measurement measured over a plurality of patient motion period amounts to each data of patient motion phases, and reconstructing E image information by each patient motion phase, respectively, by using these divided data. For example, in gated-acquisition to respiratory motion, as described in The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, minute change of breath temperature is captured, and as described in The Journal of Nuclear Medicine, Vol. 43, No. 7, pages 876 to 881, movement of the body surface of a chest part is traced with an infrared stereo camera, or the like, and thus respiration phase information is obtained, and data obtained by E measurement, based on this phase information, is divided by each of patient motion phases.
As other method for compensation of respiratory motion, there is a method for using the combined PET/CT device (refer to The Journal of Nuclear Medicine, Vol. 45, No. 8, pages 1287 to 1292). In The Journal of Nuclear Medicine, Vol. 45, No. 8, pages 1287 to 1292, attenuation compensation is executed by X-ray CT imaging in a cinema-mode, in the complex PET/CT device to obtain E image information by each respiration phase, and by using X-ray CT image at corresponding respiration phase, in preparing phase image information of each E image information by the gated-acquisition. It should be noted that, in IEEE Transactions on Medical Imaging, Vol. 18, No. 8, pages 712 to 720, non-linear superimposing technology for two image informations (Non-rigid image registration method) has been disclosed.
By execution of gated-acquisition, as described above, in E measurement, it is possible to compensate, in principle, impact accompanied with periodic patient motion. However, in order to obtain data with sufficient statistical accuracy by each patient motion phase, a long measurement time is required. Because the PET inspection over a long time gives pain to the patient (in many cases, the patient with diseases), the PET inspection is completed usually in about 20 to 30 minutes. In this case, only measurement data for several minutes or shorter, at the most, can be obtained to each patient motion phase (refer to The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219). Accordingly, E image information obtained by reconstruction of each data contains statistic noises to a great extent. In the case where filtering is executed in an imaging process to suppress the statistic noises, blurring is generated in image information. By decreasing the number of patient motion phases to be divided, statistical accuracy per unit patient motion phase can be enhanced. However, it increases patient motion within one patient motion phase time, and in the first place, it is difficult to obtain compensation effect of patient motion. That is, in a simple gated-acquisition, it cannot be executed effective patient motion compensation actually in a short time, and problems of (1) blurring of E image information and (2) decrease in quantitativeness of E image information cannot be solved basically.
In The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, it has been described that it is possible to enhance statistical accuracy, by superimposing E image information obtained by gated-acquisition on image information of a certain determined phase with non-linear deformation, and superimposing pixel values. However, in The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, there is no recognition of difficulty in deforming E image information and superimposing each other. Different from morphological image of image information to be reconstructed by X-ray CT, and T image information etc., E image information does not intend to draw a structure inside a body. Therefore, it is generally extremely difficult to deform E image information by each of patient motion phases and superimpose it each other.
It is an object of the present invention to provide an image information preparation method enabling to provide more vivid functional image information of a living body with a target of a moiety receiving impact of patient motion, in a short time.
Characteristics of the present invention to attain the above objects is an image information preparation method for preparation of superimposed first image information, by preparing first image information, in which a function of a living body was imaged at a plurality of phases obtained by dividing a certain period in terms of time; by preparing second image information, in which a structure of a living body was imaged at a plurality of phases; by determining relative displacement information among the second image information thereof by superimposing, on the second image information at a certain one the phase among a plurality of phases, the second image information at other the phases; and by superimposing, superimposed first image information on the first image information at a certain one the phase, the first image information at other the phases, by using the relative displacement information.
According to the present invention, more vivid functional image information of a living body can be obtained, in a short time, with a target of a moiety receiving impact of patient motion.
Explanation will be given below on a positron emission-type computed tomography device, which is one embodiment of the present invention, with reference to drawings, as appropriate.
As impacts given by respiratory motion on E image information, there are the following three:
The problems (1) and (2) among these have become main concerns as large factors impairing picture quality at present with dramatic enhancement of spatial resolution of the PET device. As for the problem (3), even at present with frequent experience of superimposing image information by the recently developed combined PET/CT device (refer to The Journal of Nuclear Medicine, Vol. 45, No. 1, pages 4S to 14S), solution thereof is still highly desired mainly in a field requiring specification of position of malignant tumor in good accuracy, such as in radiation therapy and biopsy.
The problem of (1) makes difficult radiation therapy planning and evaluation of therapy effect, caused by blurring of region of disease sites of internal organs and malignant tumor, or other disease moieties. In addition, even small malignant tumor with relatively not high accumulation degree can be observed vividly within displayed E image information, in the case of no patient motion present. However, in the case of patient motion present, such malignant tumor is buried in statistic noises contained in E image information, and makes difficult to recognize even presence thereof.
The problem of (2) is to underestimate the accumulation degree in general, for example, in malignant tumor with large motion. In E image information, because the accumulation degree shown in a certain pixel at a position with large patient motion, becomes time-averaged-like accumulation degree of the peripheral moieties, it provides the accumulation degree different from practical value. Therefore, the accumulation degree becomes low as shown above.
The problem of (3) is a problem occurring in superimposing E image information obtained by imaging for several minutes under natural breathing, and X-ray CT image taken in a short time under breath-holding. This superimposed image information is useful in diagnosis, however, at the peripheral of a lung field, it is observed as an image with mismatch of up to about 1 cm at the maximal. This is because of difference of handling of respiration between the PET inspection and the usual X-ray CT inspection, and thus positional correspondence is not necessarily taken. This problem has been pointed out also in the PET/CT device, and solution thereof has been desired.
By execution of gated-acquisition, as described above, in E measurement, it is possible to compensate, in principle, impact accompanied with periodic patient motion. However, in order to obtain data with sufficient statistical accuracy by each patient motion phase, a long measurement time is required. Because the PET inspection over a long time gives pain to the patient (in many cases, the patient with diseases), the PET inspection is completed usually in about 20 to 30 minutes. In this case, only measurement data for several minutes or shorter, at the most, can be obtained to each patient motion phase (refer to The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219). Accordingly, E image information obtained by reconstruction of each data contains statistic noises to a great extent. In the case where filtering is executed in an imaging process to suppress the statistic noises, blurring is generated in image information. By decreasing the number of patient motion phases to be divided, statistical accuracy per unit patient motion phase can be enhanced. However, it increases patient motion within one patient motion phase time, and in the first place, it is difficult to obtain compensation effect of patient motion. That is, in a simple gated-acquisition, it cannot be executed effective patient motion compensation actually in a short time, and problems of (1) and (2) cannot be solved basically.
In The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, it has been described that it is possible to enhance statistical accuracy, by superimposing E image information obtained by gated-acquisition on image information of a certain determined phase with non-linear deformation, and superimposing pixel values. However, in The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, there is no description on the specific method. Different from morphological image of image information to be reconstructed by X-ray CT, and T image information etc., E image information does not intend to draw a structure inside a body. Therefore, it is generally extremely difficult to deform E image information by each of patient motion phases and superimpose it each other.
Still more, as pointed out by, for example, The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, also in the case of executing gated-acquisition in E measurement, attenuation compensation is executed by using data of T measurement obtained by non-gated-acquisition in the PET device, and in the combined PET/CT device, it is executed by using X-ray CT image taken in a state of breath-holding or natural breathing. Therefore a problem is generated that between data obtained by E measurement and data obtained by T measurement, artifact is generated in attenuation compensation caused by difference of a respiration state of a patient, or deterioration of quantitativeness caused by non-execution of correct compensation. In The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, as for this problem, it has been suggested that T measurement using a γ-ray source, or gated-acquisition by X-ray CT becomes a solution. However, in The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, there is not shown a specific method for the solution.
As a method for eliminating impact of respiratory motion in E measurement, it is known imaging in a state that a patient holds breathing for a predetermined time (breath-holding imaging). This is a method for reconstruction of image by using data measured during a period when a patient holds breathing. Because a period during which a patient can hold breathing is about 30 seconds at the most, a patient is asked to hold breathing several times, if necessary. According to this method, the above-described problems (1) to (3) can be solved. However, it forces breath-holding to a patient receiving the PET inspection, and gives a tremendous pain to the patient.
Still more, to force breath-holding to a patient generates a problem in attenuation compensation. That is, because it is difficult to force breath-holding also in T measurement, E measurement is executed under breath-holding, and T measurement is executed under natural breathing. Based on this morphological difference, there are generated artifact in attenuation compensation, and deteriorated quantitativeness caused by incorrect compensation.
A method for using the complex PET/CT device, described in The Journal of Nuclear Medicine, Vol. 45, No. 8, pages 1287 to 1292, generates a problem of increased exposure to radiation, because X-ray CT imaging, which requires irradiation of high dose amount originally, is executed for a relatively long time. This method also leaves a problem of insufficient statistical accuracy described in The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, in E image information, and the problems of (1) and (2) cannot be solved basically. Still more, in the complex PET/CT device, there is time mismatching between E measurement by PET and measurement by X-ray CT. Therefore, the complex PET/CT device receives impact based on the time mismatching, such as posture change and variation of respiration of a patient in both measurements, and could generate positioning mismatch of attenuation compensation and image information.
In the PET inspection at the peripheral of a lung field, it is indispensable to eliminate impact of patient motion, in particular, respiratory motion, to obtain E image information with good quality. Accordingly, various approaches described above may be considered, however, they have not reached any solution thereof basically.
It is an object of the present invention to provide a preparation method for tomographic image information of a positron emission computed tomography device, and a positron emission computed tomography device, which enables to provide more vivid emission image information in a short time, with a target of a moiety receiving impact of patient motion.
As described above, in the PET inspection at the peripheral of a lung field, it is indispensable to eliminate impact of patient motion, in particular, respiratory motion, to obtain E image information with good quality. Accordingly, it is desirable to solve the following problems, in addition to the above problems (1) to (3):
Incidentally, as described in The Journal of Nuclear Medicine, Vol. 45, No. 2, pages 214 to 219, it has been suggested that by gated-acquisition of E image information, and by superimposing image information at each patient motion phase, on image information at a certain one patient motion phase, under deformation with non-linear strain, and adding pixel values of corresponding pixels, image information with high statistical accuracy and having respiratory motion compensated, may be obtained. If this is attained, problems (1) to (4) and (6) among the above problems can be solved. However, as described above, it is generally impossible to superimpose non-linear image information, with a clue of E image information having scarce morphological information. Accordingly, something should be worked-out to solve this. In addition, it is preferable to solve also the problem of (5).
The present inventors have studied a method for enabling to execute non-linear superimpose of E image information. Explanation will be given below on the results of this study.
In order to obtain measurement data with high statistical accuracy, by each patient motion phase of periodic patient motion (respiratory motion or heart beat) in the patient, the PET inspection over a long time is required. Therefore, it is difficult to obtain the measurement data, within a measurement period durable by the general patient without generating pain. Accordingly the present inventors have noticed on a fact that statistical accuracy can be enhanced by superimposing E image information by each patient motion phase obtained by gated-acquisition in emission measurement (hereafter called E gated-acquisition) in a shorter time, on image information of a certain one patient motion phase, and by superimposing each of pixel values. However, E image information originally has scarce morphological information, and E image information by each patient motion phase, which can be obtained, based on data obtained by gated-acquisition in a short time, has insufficient statistical accuracy. Therefore, it is difficult to execute non-linear superimpose of E image information directly among E image information by each of patient motion phases.
The present inventors have noticed, after execution of various studies, that it is good to reconstruct T image information by each patient motion phase, by execution of gated-acquisition in transmission measurement (hereafter called T gated-acquisition), in parallel to E gated-acquisition. As a result, the present inventors have found a novel method for obtaining emission superimposed image information (hereafter called E superimposed image information), by utilization of transmission superimposed image information (hereafter called T superimposed image information) obtained by non-linear superimposing of T image information themselves by each of patient motion phases, and by indirect and non-linear superimposing of a plurality of E image information themselves, by each of patient motion phases thereof. It has been found that according to this novel method, non-linear superimposing of E patient motion can be executed.
Explanation will be given below in detail on fundamental concept of the above novel method, which the present inventors have found, with reference to
T0, T1 and T2 at the upper part of
As a γ-ray source, there is used a radiation source emitting γ-ray with energy difference from energy of γ-ray generating in a patient body by positron annihilation, and a radiation detector with good energy resolution such as a semiconductor radiation detector is used. Because of discrimination, based on energy of radiation detection signals, which are output from the radiation detectors, E measurement and T measurement can be executed in parallel. Each of the discriminated data obtained by these measurements (the first packet information and the second packet information to be described later) is divided by each patient motion phase, based on patient motion phase information obtained similarly as in known gated-acquisition. By using each of the divided data, each of E image information and T image information by each patient motion phase are reconstructed. In this way, the T image information, T0, T1 and T2, and the E image information, E0, E1 and E2, can be obtained.
T gated-acquisition generally requires a very long time, although it depends on number of patient motion phases to be divided. However, by intentionally shifting relation between period of patient motion phase and rotation period of the γ-ray source, gated-acquisition in a short time is possible, also in T measurement. In a simulation by the inventors, such a result is obtained that when number of patient motion phases per one respiration period is 8, and for example, rotation period of the γ-ray source is (integer±0.1) times respiration period of a patient, T measurement data satisfying reconstruction conditions can be obtained in about 10 minutes at the most, in the case where bed position is determined at a certain one position.
Each of E image information and T image information obtained by each patient motion phase, by E gated-acquisition, and T gated-acquisition where relation between period of patient motion phase and rotation period of the γ-ray source is intentionally shifted, are image information obtained in the case where these measurements are executed in parallel (at the same time) for the same moiety (for example, a moiety receiving impact of patient motion) of the same patient. There is no essential morphological difference of the patient, between the E measurement and the T measurement executed in parallel. That is spatial relative displacement between two E image information corresponding to two patient motion phases in the E gated-acquisition, become the same as that between two T image information corresponding to these patient motion phases in the T gated-acquisition. For example, spatial relative displacement [F10] between E image information E0 at patient motion phase 0 (reference phase), and E image information E1 at patient motion phase 1, in the E gated-acquisition is the same as the deformation between T image information T0 of the reference phase and T image information T1 at the patient motion phase 1. Accordingly, by determination of information of the relative displacement between two patient motion phases in T gated-acquisition (for example, [F10] or [F20]), and by applying the information of this relative displacement to two E image information at the two patient motion phases, non-linear superimpose of E image information at one patient motion phase, on E image information at the other patient motion phase (for example, reference phase) is possible. As a result, emission superimposed image information with non-linear superimpose can be obtained.
It should be noted that because T image information enables to acquire morphological information of such as a body contour, a lung field and a bone, by using non-linear superimpose technology described in, for example, IEEE Transactions on Medical Imaging, Vol. 18, No. 8, pages 712 to 720, it is possible to superimpose T image information at each patient motion phase, on image of a reference phase, under non-linear strain.
T
jo
=[F
jo
]·T
j(j=1,2) (1)
wherein [Fjo] is a conversion matrix representing relative displacement from T image information Tj corresponding to a respiration phase j, to reference image information T0. Tj0 at the left-hand side is T image information obtained by mapping T image information Tj at a phase j, on morphology at a patient body of a reference phase (reference image information T0). Each of these a plurality of T image information is expressed as a vector. It should be noted that it is easy to extract and store spatial mapping as a mapping function from a superimposing process.
Superimpose (mapping) of E image information (having ambiguous morphology) Ej corresponding to a respiration phase j on reference image information E0 is executed by using the conversion matrix Fj0 obtained from superimpose of the above T image information on E image information Ej. E image information Ej, which is superimposed on reference image information E0, is a mapping represented by the equation (2). Reference image information E0 is E image information corresponding to a reference phase.
E
jo
=[F
jo
]·E
j(j=1,2) (2)
As described above, after superimposing each of E image information at all patient motion phases other than a reference phase, on reference E image information at a reference phase, each pixel value of each E image information is added to pixel value of reference E image information. By this processing, E image information E0′ with high statistical accuracy, for patient morphology at a reference phase, can be obtained. This is expressed by the equation (3).
E
0′=(E0+E10+E20)/3 (3)
E10 of the equation (3) represents calculated value when j in the equation (2) is 1, and E20 represents calculated value when j in the equation (2) is 2.
By non-linear superimpose of corresponding E image information at a patient motion phase, based on relative displacement between corresponding T image information at two patient motion phases as described above, problems other than the above-described problem (5) can be improved. The problem (5), that is a problem caused by morphological mismatching in attenuation compensation, can be solved by projection calculation based on these T image information, or by attenuation compensation using measurement data (projection data) used in reconstruction of T image information, because T image information having coincident morphology as E image information by each patient motion phase can be obtained.
The above-described respiratory motion compensation method (a patient motion compensation method) is calculated and processed by a patient motion compensation device 38 of
Explanation will be given on a positron emission computed tomography device 1 (the PET device), which is one suitable embodiment of the present invention, with reference to
The computed tomography device 2 has a housing 45 (refer to
Each detector 9 uses cadmium telluride (CdTe) as a semiconductor element part excellent in energy resolution. It is also possible to use cadmium zinc telluride (CZT), lead iodide (PbI2), thallium bromide (TlBr) and gallium arsenide (GaAs), as the semiconductor element part. Each detector 9 detects γ-ray of 511 keV (the first γ-ray) emitted from the patient 29 caused by the PET pharmaceutical, and γ-ray (the second γ-ray) emitted from a γ-ray source (refer to
The cover member 19a, which is a part of the 19, is attached at one end part of the housing member 19 detachably. A unit joint FPGA (Field Programmable Gate Array, hereafter called a FPGA 17), and connectors 18, 24, 25 and C2 are installed at the cover member 19a. Signal wirings (not shown) of a plurality of connectors C2 are connected to the FPGA 17, and power source wirings (not shown) of these connectors C2 are connected to the connector 25. A signal wiring (not shown) of the connector 18 is connected to the FPGA 17, and a power source wiring (not shown) of the connector 18 is connected to the connector 24. The voltage adjustment device 21 is connected to the connector 24 and the connector 25.
One analog ASIC 10 is provided with a plurality of analog signal processing circuits (analog signal processing devices) 10a. The analog signal processing circuit 10a having a fast system and a slow system is installed by each detector 9. The fast system has a timing pick-off circuit 10b for outputting a timing signal for specifying detection time of γ-ray. The slow system is connected with and installed with a polar amplifier (a linear amplifier) 10d, a band-pass-filter (a waveform fixing device) 10e, and a peak-hold circuit (wave height value holding device) 10f, in this order. The analog signal processing circuit 10a has a charge amplifier (pre amplifier) 10c connected to the timing pick-off circuit 10b and the polar amplifier 10d. The charge amplifier 10c is connected to one detector 9.
A digital ASIC 12 has, as shown in
The detector 9, which detected incident γ-ray, outputs a γ-ray detection signal. The detector 9 outputs the first γ-ray detection signal, when the γ-ray is the first γ-ray, and outputs the second γ-ray detection signal, when the γ-ray is the second γ-ray. In the case of describing hereafter simply as γ-ray without differentiating the first γ-ray and the second γ-ray, it means both of the γ-rays, and in the case of describing simply as a γ-ray detection signal without differentiating the first γ-ray detection signal and the second γ-ray detection signal, it means both of the γ-ray detection signals. The γ-ray detection signal, which is output from the detector 9, is amplified with the charge amplifier 10c and the polar amplifier 10d. The amplified γ-ray detection signal is input to the peak-hold circuit 10f via the band-pass-filter 10e. The peak-hold circuit 10f holds wave height value of the γ-ray detection signal.
The wave height value signal, which is output from the peak-hold circuit 10f, is input to a packet data generator 15 by being converted to a digital signal with the ADC 11. The timing pick-off circuit 10b receives input of the γ-ray detection signal amplified by the charge amplifier 10c, and outputs a timing signal showing detection timing of γ-ray. This timing signal is input to the corresponding time determination circuit 14. The time determination circuit 14 determines detection time of γ-ray, based on a clock signal in receiving an input timing signal, and outputs detection time information.
The packet data generator 15, when it receives input information of detection timing, specifies detector ID of the detector 9 corresponding to the time determination circuit 14, which output the detection time information. The packet data generator 15 generates packet information, which is digital information containing detection time information, detector ID information (detector position information) and wave height value information (energy information of the γ-ray detection signal), corresponding to one detector 9. This packet information is input to the data transfer circuit 16. A signal wiring (not shown) connected to the data transfer circuit 16 is connected to a signal wiring connected to the connector C2 via the connector C1. It should be noted that the electric power wiring connected to the connector C2 is connected to the electric source wiring (not shown) connected to the connector C1. The latter electric power wiring is connected to each element installed at the module substrate 7 such as detector 9 and the timing pick-off circuit 10b, arranged inside the substrate member 8.
The data transfer circuit 16 of each module substrate 7 arranged inside one detector unit 6 outputs packet information to the FPGA 17 of the detector unit 6. Each packet information, which is output from each FPGA 17 of all detector units 6 installed at the computed tomography device 2, is sent to a signal discrimination device 31 of the data processing device 30 via an information transfer wiring (not shown) connected to each connector 18.
The radiation source rotation device 54 has a rotation drive device 55 (for example, a motor), the rotation member 56, a ring-like gear member 57, an axis support member 58 and a rolling member 59. The rotation member 56 is arranged at the rear end part of the housing 45. A part of the rotation member 56 is arranged between the housing 45 and the rear surface end shield 60, and is attached at the rolling member 59 (for example, a thrust bearing) attached at the housing 45. The rolling member 59 supports the rotation member 56. The rotation drive device 55 is attached at the housing 45. The rolling member 59 supports the rotation member 56. The rotation drive device 55 is attached at the housing 45. A gear installed at the rotation axis of the rotation drive device 55 is engaged with the gear member 57 installed at the rotation member 56 encompassing a circumference part of the rotation member 56. The rear surface end shield 60 rotates with the rotation member 56. An encoder 63 attached at the rotation drive device 55 is jointed to the rotation axis of the rotation drive device 55.
A radiation source device 48 has a radiation source 26, which is a γ-ray source, a radiation source support axis 49 and a holding member 50. The radiation source 26 is attached at one end of the radiation source support axis 49, and the holding member 50 is attached at the other end of the radiation source support axis 49. The radiation source support axis 49 is arranged so as to be parallel to the center axis (the center axis of the measurement space R) Z of the computed tomography device 2. The radiation source 26 is arranged at nearer position to the Z side than the detector unit 6, that is, between the detector unit 6 and the bed 27. The radiation source support axis 49 penetrates the axis support member 58 installed at the rotation member 56. The axis support member 58 is a radiation shielding body and shields one end part of the radiation source storage chamber 61 formed inside the rotation member 56. A notch part 60a for the radiation source 26, stored in the radiation source storage chamber 61, to pass through is installed at the rear surface end shield 60. Radiation source shields 64 and 65, which are radiation shielding bodies, are installed. The radiation source shield 64 is arranged facing to the notch part 60a, and the radiation source shield 65 is arranged facing to the position of the radiation source storage chamber 61.
The radiation source 26 is provided with a radio isotope, which emits the second γ-ray with energy different from that of the first γ-ray. As this radio isotope, cesium 137 emitting γ-ray of 662 keV is used. Instead of cesium 137, cobalt 57 (emitting the second γ-ray of 122 keV), technetium 99m (emitting the second γ-ray of 140 keV), tellurium 123m (emitting the second γ-ray of 159 keV), cerium 139 (emitting the second γ-ray of 166 keV), gadolinium 153 (emitting the second γ-ray of 153 keV) and americium 241 (emitting the second γ-ray of 57 keV) etc. may also be used. An X-ray source etc. as other radiation source, may also be used.
The radiation source transitory transfer device 53 is provided with a transfer device 51 and a guide member 52. The guide member 52 extends in an axial direction of the computed tomography device 2, and attached at the side surface of the rear end of the housing 45. The transfer device 51 moves along the guide member 52 in the axis direction thereof, and has a channel 62 inserted with the holding member 50. The transfer device 51 enables to move in the axis direction thereof, in a state that the radiation source 26 is present at the lowest position, that is, in a state that the holding member 50 is inside the channel 62. By movement of the transfer device 51, the radiation source 26 is moved between the front surface end shield 5 and the axis support member 58 in the axis direction of the computed tomography device 2. The radiation source 26 is stored in the radiation source storage chamber, when T measurement is not executed. When T measurement is executed, the channel 62 is drawn out from the holding member 50, by driving the rotation drive device 55 and rotating the rotation member 56, and the radiation source 26 is rotated around the patient 29 lying on the bed 27.
The signal discrimination device 31 is connected to each connector 18 of each detector unit 6, and still more connected to each of the coincidence counting device 32 and the second radiation processing device 34. The radiation source position detection device 41 is connected to each of the encoder 63, the second radiation processing device 34 and the radiation source rotation control device 42. The radiation source rotation control device 42 controls drive of the rotation drive device 55. The memory device 35 is connected to each of the coincidence counting device 32, the second radiation processing device 34, the patient motion phase information acquiring device 43, the phase information generation device 44, the first tomographic image preparation device 36, the second tomographic image preparation device 37, the relative displacement information generation device 39, the tomographic image superimposing device 40 and an information output device 66. The relative displacement information generation device 39 is connected to the tomographic image superimposing device 40. The patient motion phase information acquiring device 43 is connected to a respiration monitor device 28.
The respiration monitor device 28 is a device for monitoring breath temperature of the patient 29, and is mounted on the face of the patient 29. The patient motion phase information acquiring device 43 receives input of breath temperature information measured with the respiration monitor device 28, and determines interval, as respiration period, between change timings from exhalation to inhalation, and between peaks when breath temperature becomes highest. As the respiration monitor device, there may be used an infrared camera or an optical stereo camera, for measuring displacement of the chest part surface skin of the patient 29. In using the infrared camera or the optical stereo camera, the relevant camera, which is installed on a supporting member such as a tripod, is arranged at a position making possible photographing of the patient 29 at the vicinity of the computed tomography device 2. In this case, the patient motion phase information acquiring device 43 analyzes displacement waveform of the chest part surface skin, based on picture information, which is input from the infrared camera or the optical stereo camera, and determines respiration period. Alternatively, a device enabling respiration monitor is separately prepared, and a signal, which is output from this device and gives a clue of phase information analysis, may be input to the patient motion phase information acquiring device 43.
Explanation will be given in detail on the PET inspection using the PET device of the present embodiment, and on preparation of image information, based on information obtained by this inspection.
An operator (a doctor or a radiation engineer) inputs information necessary to the PET inspection from an input device (not shown) installed at an operator console (not shown). This input information contains information relating to the patient 29 and the PET pharmaceutical, and still more information for respiratory motion compensation. For example, information for respiratory motion compensation includes 12 minutes, which is PET inspection time, and 8, which is number of patient motion phases per one respiration phase. The operator still more inputs an inspection start order from the input device. A bed drive device (not shown), which received input of this inspection start order moves the bed 27, on which the patient 29 is lying, who was administered with the PET pharmaceutical, in a longitudinal direction thereof, and inserts the patient 29 inside the measurement space R. The patient 29 is positioned at a predetermined position of the axis direction inside the measurement space R.
The respiration monitor device 28, which received input of this inspection start order, starts monitoring of respiration of the patient 29. The patient motion phase information acquiring device 43 determines respiration period, as described above, based on measurement information of the respiration monitor device 28, that is, temperature measurement value of breath of the patient 29. The patient motion phase information acquiring device 43 orders a memory device 35 to memorize respiration waveform information of each respiration period determined. The information output device 66 outputs the respiration waveform information by each respiration period of the patient 29, which was read out from the memory device 35, onto the display device 33. The output of relating information onto the display device 33 by the information output device 66 is executed by a request order from the operator, which is input from the above input device. When the respiration waveform information of the patient 29, which is displayed on the display device 33, is stabilized, the operator inputs T measurement start order to the input device so as to start T measurement. The respiration waveform information stabilizes in several minutes. The T measurement is executed in parallel to E measurement within a period where E measurement is executed.
The T measurement start order is input from the input device to each of the phase information generation devices 44, the radiation source linear transfer control device (not shown) and the radiation source rotation control device 42.
The phase information generation device 44, which received input of the T measurement start order, divides the respiration waveform information of each respiration period by a time axis, based on the respiration waveform information by each respiration period, which is input from the patient motion phase information acquiring device 43, and number of patient motion phases (for example, 8) in one respiration period, which is input by the operator from the above input device. The phase information generation device 44 provides time information by each divided patient motion phase. In providing this time information, start time of respiration period and detection time of γ-ray, which is determined by the time determination circuit 14, may be mismatched unless being synchronized, and gives a trouble in preparation of E image information and T image information at each patient motion phase, by the first tomographic image preparation device 36 and second tomographic image preparation device 37. Explanation will be given on a method for synchronizing E gated-acquisition and T gated-acquisition. A packet containing time information is sent from the detector unit 6 to the data processing device 30. The patient motion phase information acquiring device 43 acquires the packet containing time information in real time, via the memory device 35. The phase information generation device 44 obtains time information from the patient motion phase information acquiring device 43. For example, in the case where respiration period determined is 4 seconds, because patient motion phase function thereof is 8, which was input from the input device, time width of one patient motion phase becomes 0.5 second. Time information is provided by each of 8 patient motion phases, divided by each 0.5 second. This processing is repeated by each respiration period. In addition to uniform division of respiration period, in the case where displacement of an up-down motion of the chest part skin surface of the patient 29 was determined, based on picture of an infrared camera, patient motion phase may be divided by using indices of waveform and displacement amount of the displacement of the up-down motion. Information on each patient motion phase, provided with time information by each respiration period, is memorized in the memory device 35.
A control device for controlling the radiation source linear transfer device 53, which received input of T measurement start order, outputs a drive order to the transfer device 51. The radiation source support bar 49 engaged with the transfer device 51 moves toward the front surface end shield 5. By this movement, the radiation source 26 moves from the radiation source storage chamber 61 to the outside, and is set at predetermined position between the front surface end shield 5 and the rear surface end shield 60.
The radiation source rotation control device 42 determines a rotation period and a rotation number of the radiation source 26 suitable to execute T gated-acquisition in a short time, by using information on respiration period and information on each of divided patient motion phases, which are memorized in the memory device 35. The radiation source rotation control device 42, which received input of T measurement start order, controls rotation of the rotation drive device 55 so as to attain the rotation period and the rotation number determined, after the radiation source 26 is set at the above predetermined position. For example, in the case where respiration period of a patient is 4 seconds, the radiation source 26 is rotated around the patient 29 for about 12 minutes, in a period of 16.4 seconds, that is 4.1 times thereof. Rotation force of the rotation drive device 55 is conveyed to the rotation member 56 via the gear member 57 to rotate the rotation member 56. The radiation source 26 rotates with the rotation member 56, and rotates around the patient 29. The second γ-ray of 662 keV emitted from the radiation source 26 is irradiated onto the patient 29 on the bed 27 from the circumference, accompanying with rotation of the radiation source 26. T measurement is executed under rotation of the radiation source 26.
For example, when rotation of the radiation source 26 is started, the operator inputs a data collection start order to the above input device. When this data collection start order is input to the signal discrimination device 31, the signal discrimination device 31 starts input of each packet information, which was output from each detector unit 6.
Explanation will be given on generation of packet information in each of E measurement and T measurement, which is input to the signal discrimination device 31.
During a period when the patient 29 is inserted in the measurement space R, all of the detectors 9 encompass the patient 29. In this state, E measurement is executed. Pair γ-rays (a pair of the first γ-rays) generating in annihilation of a positron, which was generated caused by the PET pharmaceutical accumulated at an affected part of malignant tumor, is injected into a pair of the detectors 9, which are located in about 180 degree opposite direction of the computed tomography device 2, and detected by these detectors 9. The detector 9, which has detected the first γ-ray, outputs the first γ-ray detection signal. The timing pick-off circuit 10b of the analog signal processing circuit 10a, which receives input of this first γ-ray detection signal, outputs a timing signal, and the peak-hold circuit 10f outputs a wave height value signal. The time determination circuit 14, which receives input of this timing signal, generates detection time information of the first γ-ray, which is determined based on the timing signal, as described above. The packet data generator 15, which received input of the wave height value information and the detection time information, which were converted to digital signals by ADC 11, generates packet information for the detected first γ-ray (hereafter called the first packet information). This first packet information contains detection time information, detector ID information and wave height value information, for the first γ-ray. The first packet information obtained by E measurement is input to the signal discrimination device 31.
In T measurement, the second γ-ray emitted from the radiation source 26 and transmitted through the patient 29 is detected with the detector 9. The second γ-ray detection signal, output from the detector 9, which detected the second γ-ray, is processed by the analog signal processing circuit 10a and the packet data generation device 13, similarly as in the first γ-ray detection signal. The packet data generator 15 generates packet information for the detected second γ-ray (hereafter called the second packet information). This second packet information contains detection time information, detector ID information and wave height value information for the second γ-ray. The second packet information obtained by E measurement is also input to the signal discrimination device 31.
In terminating T measurement, the operator inputs a T measurement termination order from the above input device. The radiation source rotation control device 42, which received input of the T measurement termination order, outputs a stop control order to the rotation drive device 55, and stops the rotation drive device 55, when the radiation source 26, which is a γ-ray source, reached the lowest position. In this state, the holding member 50 is located inside the channel 62 of the transfer device 51. The radiation source linear transfer control device, which received input of the T measurement termination order, controls the transfer device 51, so as to move apart from the housing 45. Because the transfer device 51 moves so as to be apart from the housing 45, the radiation source 26, which is a γ-ray source, is stored inside the radiation source storage chamber 61, and irradiation of the second γ-ray onto the patient 29 is stopped.
After input of the data collection start order, the signal discrimination device 31 receives both inputs of the first packet information and the second packet information, which were output from each detector unit 6, because E measurement and T measurement are executed in parallel, and discriminates these packet information, based on energy of detected γ-ray, that is wave height value information. The signal discrimination device 31 outputs the first packet information, containing the wave height value information corresponding to energy of the first γ-ray, that is, the wave height value information in a range corresponding to an energy of, for example, 450 to 550 keV, to the coincidence counting device 32. The second packet information, containing the wave height value information corresponding to energy of the second γ-ray, that is, the wave height value information in a range corresponding to an energy of, for example, 570 to 650 keV, is output to the second radiation processing device 34. Exchange of output destination of these packet information is executed by exchanging an exchange switch (not shown) installed inside the signal discrimination device 31.
The coincidence counting device 32 executes coincidence counting by using detection time information and detector ID, obtained based on the first γ-ray detection signal, and specifies a pair of the detectors 9, which detected a pair of the first γ-rays generated by annihilation of one positron. The coincidence counting device 32 orders the memory device 35 to memorize each detector ID information and detection time information contained in the first packet information of a pair of the specified detectors 9. Each detector ID information and detection time information, obtained by coincidence counting, are called the first detection information.
The radiation source position detection device 41 receives input of an output signal (rotation angle information of the rotation drive device 55) of the encoder 63, and determines position information of the radiation source 26, which is rotating, based on this output signal (hereafter called radiation source position information). The radiation source position information is input to the radiation source rotation control device 42, as the second radiation processing device 34 and feed-back information.
The second radiation processing device 34 orders the memory device 35 to memorize detector ID information and detection time information of the detector 9, which detected the second γ-ray, contained in the second packet information, and additional radiation source position information. The detector ID information, detection time information, and radiation source position information, which are output from the second radiation processing device 34, are called the second detection information. Because the first detection information and the second detection information each contains detection time information, it contributes to correspond to patient motion phase information provided with time information, which is generated by the phase information generation device 44.
The second tomographic image preparation device 37 reconstructs T image information by each patient motion phase, by using the second detection information memorized in the memory device 35, and patient motion phase information provided with time information. Specifically, the second tomographic image preparation device 37 prepares T image information for the above certain patient motion phase, by using time information at a certain patient motion phase, that is the second detection information having time information contained in between time information of this patient motion phase and time information of the next patient motion phase. This T image information is provided with corresponding patient motion phase time information, and memorized by the memory device 35. In this way, T image information by each of patient motion phases, 0, 1, 2, - - - 7, for example, T image information, T0, T1, T2, - - - T7 (refer to
The first tomographic image preparation device 36 reconstructs E image information by each patient motion phase, by using the first detection information memorized in the memory device 35, and patient motion phase information provided with time information. E image information for the above certain patient motion phase is prepared, by using the first detection information having time information contained in between time information of this patient motion phase and time information of the next patient motion phase. This E image information is provided with corresponding patient motion phase time information. This E image information is provided with time information of corresponding patient motion phase, and memorized by the memory device 35. In this way, information by each of patient motion phases, 0, 1, 2, - - - 7, for example, E image information, E0, E1, E2, - - - E7 (refer to
In reconstruction of E image information at a certain patient motion phase, attenuation compensation for the first detection information is executed, based on the second detection information at that patient motion phase, or tomographic image relating to a transmittance of the second γ-ray, reconstructed at that patient motion phase, and E image information is reconstructed. Here, tomographic image relating to the transmittance of the second γ-ray corresponds to T image information at a patient motion phase.
The first tomographic image preparation device 36 and the second tomographic image preparation device 37 prepare E image information and T image information by each patient motion phase, respectively, by using tomographic image reconstruction algorism, for example, a filtered-back-projection method etc. Each of E image information and T image information is tomographic image information, and is prepared by each 8 sets by each respiration period. The first tomographic image preparation device 36 executes attenuation compensation. In the case where attenuation compensation is necessary, it is executed from T image information at the same patient motion phase for the first detection information, however, in the present embodiment, by parallel measurement (substantially coincidence counting) of T measurement and T measurement, morphological identity of the patient 29 is guaranteed. Therefore, in the present embodiment, artifact and deterioration of quantitativeness accompanied with position mismatch in E gated-acquisition and T gated-acquisition in attenuation compensation, are far less as compared with conventional E gated-acquisition.
It should be noted that each E image information by each patient motion phase obtained at the stage before preparation of superimposed image information, based on E gated-acquisition, corresponds to measurement data at a divided period of total inspection time with number of patient motion phases. That is, when patient motion phases are divided equally with time, it is one reconstructed from measurement data of 8 division of 12 minutes, that is 1.5 minute per one patient motion phase. In usual administration amount of the PET pharmaceutical, image information can be obtained, however, statistical accuracy is absolutely insufficient, and thus it is image information containing many noises. In conventional gated-acquisition, because of absence of a preparation step for superimposed image information, based on E gated-acquisition, to be described later, it is obtained only 8 image reconstructed from measurement data of substantially 1.5 minute. To obtain clear image with low noise, it is generally required measurement of equal to or longer than 3 minutes, therefore, clear image cannot be obtained by conventional gated-acquisition.
Explanation will be given on specific processing for non-linear superimposing of T image information themselves and E image information themselves by each patient motion phase, described by using
The tomographic image superimposing device 40 receives input of each E image information provided with time information at a patient motion phase memorized in the memory device 35 (for example, in the case where 8 patient motion phases are present in one respiration period, E image information, E0, E1, E2, - - - , E7) and the above-described each patient motion phase information provided with time information. The tomographic image superimposing device 40 superimposes E image information, E1, E2, - - - , E7, at each of patient motion phases 1 to 7 in a certain one respiration period, on E image information E0 at a patient motion phase 0 (a reference phase) non-linearly, by using a plurality of conversion matrix information [F10], [F20], - - - , [F70], to prepare E superimposed image information for one respiration period. This non-linear superimposing processing of image information is executed by using the above-described algorism called a “Non-rigid Image Registration Method”. The tomographic image superimposing device 40 executes addition processing of pixel values of each pixel at the same spatial position, for all the pixels of E image information at reference phase, by using each E image information at each patient motion phase. The resulting E superimposed image information is memorized in the memory device 35. The above-described preparation process of E superimposed image information is executed by each respiration period within a period of the PET inspection.
The tomographic image superimposing device 40 still more executes addition processing of pixel values of each pixel at the same spatial position of E superimposed image information, for all the pixels of certain one E superimposed image information, by using each E superimposed image information at reference phase prepared at each respiration period. Statistical-noise-suppressed motion-compensated E image information, obtained by such addition of pixel values, becomes more vivid image information with high statistical accuracy. The statistical-noise-suppressed motion-compensated E image information is memorized in the memory device 35.
The information output device 66 reads out, from the memory device 35, each T image information, and each E image information by each patient motion phase, each T superimposed image information, and each E superimposed image information by each respiratory period, and statistical-noise-suppressed motion-compensated T information, and statistical-noise-suppressed motion-compensated E information, each singly by image information, or together of a plurality kinds of image information, and outputs to the display device 33. The display device 33 displays image information, which was input. Image information displayed at the display device 33 is read out from the display device 33 by the information output device 66, based on an image information display order, which the operator inputs from the input device of the operator console.
In the present embodiment, by using relative displacement information, that is, conversion matrix information, obtained by superimposing, on T image information of a certain patient motion phase (for example, a reference phase), T image information of other patient motion phases, there is superimposed, on E image information at this one patient motion phase, E image information at the above other patient motion phases, therefore E image information at each of patient motion phases in one respiration period can be superimposed all together. Accordingly, more vivid E image information (statistical-noise-suppressed motion-compensated image), with a target of a moiety receiving impact of patient motion (for example, respiratory motion), can be obtained in a shorter time. According to the more vivid E image information (statistical-noise-suppressed motion-compensated image), obtained by the present embodiment, diagnosis of malignant tumor, present at a moiety receiving impact of patient motion, can be executed in good accuracy. Statistical accuracy, expected to this E image information in the present embodiment, corresponds to that in E measurement of 12 minutes. Usually, by execution of E measurement for 12 minutes, statistical accuracy can be satisfied sufficiently. In addition, this statistical-noise-suppressed motion-compensated image information is vivid image with small blurring and high quantitativeness, because respiratory motion is also compensated by each of T image information at each of patient motion phases in the present embodiment. In addition, because of no generation of position mismatch of the patient 29, in T measurement and E measurement in attenuation compensation, artifact accompanied with the position mismatch is also not generated. The patient 29 can continue respiration in a calm state, and is not required to hold breath during the period of the PET inspection in the present embodiment.
As described above, reason for enabling to utilize relative displacement information obtained, based on T image information at a patient motion phase, in superimposing E image information at different patient motion phases is that E measurement and T measurement are executed in parallel, and the reason owes to detection of the first γ-ray and the second γ-ray with one detector 9. In many cases, one detector 9 outputs the first γ-ray detection signal and the second γ-ray detection signal with time mismatch. The first packet information and the second packet information prepared based on these γ-ray detection signals, can be discriminated easily by the signal discrimination device 31, based on energy of the γ-ray detection signals. That is, because the signal discrimination device 31 receives input of wave height value corresponding to energy of the γ-ray detection signal obtained by the peak-hold circuit (a wave height value generation device) 10f, the first packet information and the second packet information obtained by the peak-hold circuit 10f for the determination, can be classified easily.
If necessary, in adding pixel values of set of E image information at other patient motion phases, to pixel value of E image information at reference phase, by execution of the addition processing, by adjusting the pixel values in consideration of stretching and contracting degree of the E image information, which is deformed non-linearly, quantitativeness of accumulation degree of the PET pharmaceutical can be enhanced still more.
On one or a plurality of display device 33 installed at the operator console, E image information and T image information at the same patient motion phase, or T superimposed image information and E superimposed image information at a reference phase can be displayed, along with statistical-noise-suppressed motion-compensated E image information. By such display, the statistical-noise-suppressed motion-compensated E image information can be evaluated.
As described above, in order to obtain more vivid emission image information in a short time, with a moiety receiving impact of patient motion as a target, there is executed a tomographic image information preparation method of a positron emission computed tomography device, wherein first image information is prepared at a plurality of phases obtained by dividing a certain period, in terms of time, based on a plurality of first information obtained by a plurality of first radiation detection signals, which are output from a plurality of radiation detectors surrounding a bed, when first radiation caused by a radiopharmaceutical is input; second tomographic image information is prepared at a plurality of phases, based on a plurality of second information obtained by a plurality of second radiation detection signals, which are output from a plurality of radiation detectors, when second radiation emitted from a radiation source is input; relative displacement information among a plurality of second tomographic image information is determined by superimposing, on the second tomographic image information at a certain one phase among a plurality of phases, a plurality of second tomographic image information at other phases; and superimposed first tomographic image information is prepared by superimposing, on the first tomographic image information at a certain one phase, a plurality of first tomographic image information at other phases, by using relative displacement information.
Description was given above on processing of superimposing, on T image information T0 at a patient motion phase 0, T image information, T1, T2, - - - , T7, of other patient motion phases. However, the relative displacement information generation device 39 may execute similar superimposing processing of T image information and addition processing of pixel values, also in superimposing of each of T image information, T0, T2, - - - , T7, at for example, patient motion phases, 0, 2, 3, - - - , 7, on T image information of the patient motion phase 1, and addition of pixel values. Superimposing of other T image information on each of the residual T image information, T2, T3, - - - , T7, and addition of pixel values are also executed similarly. T superimposed image information at each of patient motion phases, 0, 2, 3, - - - , 7, provide image information with high statistical accuracy. By execution of attenuation compensation of E image information at corresponding patient motion phase, by using these T superimposed image information, propagation of statistical noises contained in T image information in the attenuation compensation process to E image information can also be suppressed.
Still more, by interpolation in a time direction, by using both E image informations of adjacent patient motion phases, vivid E image information for an arbitrary respiration phase can be obtained. By matching an arbitrary respiration phase to that of breath-holding X-ray CT image of the same patient, which is imaged at a separate time, there can also be solved a problem of position mismatching between PET image and CT image, caused by difference of a respiration state in E image information and X-ray CT image information. However, in this case, it is necessary to match a bed shape or a patient posture, in advance.
The above-described embodiment used, as a target, gated-acquisition corresponding to respiratory motion, however, it is applicable also to gated-acquisition corresponding to heart beat.
It should be noted that in a nuclear medicine diagnosis device, there may be the case where energy is provided to a plurality of radiation detectors, due to scattering of γ-ray from the patient inside a certain radiation detector, and absorbing by other radiation detectors. For such a case, it is considered to judge whether γ-ray before being scattered is γ-ray from the radiopharmaceutical administered to the patient P, based on radiation detection information by two or more radiation detectors, and if so, it is processed as an effective signal. Hereafter, such a method is called scattered ray processing. The nuclear medicine diagnosis device has a scattered light processing unit for specifying a plurality of radiation signals by radiation scattered by the radiation detector, as one radiation signal, based on output signals output from the radiation detectors. Here, the scattered light processing unit may be installed at the more upstream side of data processing than the signal discrimination device 31 of the 30. Because effective signals increase by the scattered light processing, diagnosis image with good accuracy can be expected.
It should be noted that, in the above embodiments, an example of a positron emission computed tomography device was shown, however, it may be an SPECT device, as long as it is capable of imaging functional image of a living body. In addition, it is applicable to not only two-dimensional tomographic image but also relation of three-dimensional functional image and structural image of a living body similarly. The above-described preparation method for image information may be applicable to image taken by computer, where medical image receives impact of patient motion. That is, to obtain more vivid functional image information of a living body in a short time, with a moiety receiving impact of patient motion, as a target, there is executed an image information preparation method of an image processing device, wherein first image information is prepared, with function of a living body imaged at a plurality of phases obtained by dividing a certain period, in terms of time; second image information is prepared, with structure of a living body imaged at a plurality of phases; relative displacement information among the a plurality of second image information is determined by superimposing, on the second image information at a certain one the phase among a plurality of phases, a plurality of second image information at other phases; and superimposed first image information is prepared by superimposing, on the first image information at a certain one the phase, a plurality of first image information at other phases, by using the relative displacement information.
It should be noted that in the above embodiments, a certain period on E image information and T image information was divided into 3 or 8, however, it may be divided into any specified number. In addition, an example of equal division of period was shown, however, it may not be equal division, and time of a certain division area may be set shorter or longer than time of other division areas. For example, by setting time shorter for an area receiving larger impact of patient motion, and setting time longer for an area receiving smaller impact of patient motion, image processing can be executed by differentiating the area receiving smaller impact of patient motion, and the area receiving larger impact of patient motion, and more vivid image can be obtained. In the case of setting time shorter or longer for a certain division area as compared with other division area, by dividing with imaging time by added amount, instead of dividing with superimposed image number, 3, in the right-hand side of the equation (3), superimposed image can be obtained.
As for divided areas, it is not necessary to use all of the areas in superimpose, and a part of the areas may be specified and used in superimpose. For example, by specifying and superimposing an area including a time for complete air exhalation, which is an area with small impact of patient motion, and an area including a time for complete air inhalation, among all areas, image with small impact of patient motion can be obtained and quick measurement is possible. In addition, by elimination of image area with large impact of patient motion, and with index expressing a blurring degree over predetermined value, image accuracy can be enhanced, while putting importance on quick measurement.
It should be noted that in the above embodiments, an example was given to take synchronization of patient motion period by imaging E image information, which is a functional image of a living body, and T image information, which is a structural image of a living body, by one imaging device, however, it is also possible to take synchronization of patient motion period of a functional image of a living body, taken by a certain imaging device, and a structural image of a living body, taken by a separate imaging device. For example, by specifying period of the functional image and period of the structural image, and by assigning areas to be divided so that each of the starting points and the lengths of the periods are matched, synchronization of the functional image and the structural image can be taken.
It should be further understood by those skilled in the art that although the foregoing description has been made on embodiments of the invention, the invention is not limited thereto and various changes and modifications may be made without departing from the spirit of the invention and the scope of the appended claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2008-081500 | Mar 2008 | JP | national |
