Nuclear magnetic resonance imaging device, nad body motion information processing method

Abstract

To provide a nuclear magnetic resonance imaging device and a body motion information processing method which can detect and classify a body motion of a subject.SOLUTION: A nuclear magnetic resonance imaging device includes: an imaging part for acquiring an image of a subject by measuring a nuclear magnetic resonance signal generated by the subject; and a body motion information processing part having a processor for processing motion information of the subject provided to the imaging device. The body motion information processing part includes a body motion information calculating part and a body motion information classifying part. The processor receives the signal from the measurement device for measuring the motion information of the subject, calculates the body motion information based on the signal, and classifies the body motion information.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an apparatus for processing information relating to the movement of a subject under examination. [Background technology]

[0002] In an examination using a medical imaging device such as a magnetic resonance imaging device (hereinafter referred to as an MRI (Magnetic Resonance Imaging) device), the subject (patient) is examined while lying on a table or loosely fixed. If the subject moves during the examination, artifacts will appear in the images obtained by the medical imaging device, hindering diagnosis. Furthermore, if the subject moves suddenly or significantly, or if the subject moves significantly from the examination position, or even falls off the bed, the examination must be interrupted. However, during the examination, the technician or doctor (referred to as a user) is often gazing at the monitor displaying the image and may not notice.

[0003] 2. Description of the Related Art Conventionally, as a technique relating to medical imaging apparatuses, a technique for reducing the influence of the movement of a subject during an examination on a diagnostic image has been developed and proposed.

[0004] For example, Patent Document 1 proposes that in an MRI apparatus, body movement is detected by a camera attached in or near the examination space, and that scanning is stopped or restarted when body movement is detected. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Laid-Open No. 2006-346235 Summary of the Invention [Problem to be solved by the invention]

[0006] Incidentally, body movements include large body movements that affect safety and small body movements that affect image quality. For example, if the body movement is large and affects safety, the user needs to go to the subject and provide immediate support. On the other hand, if the body movement is small and affects image quality, the user can continue imaging. In other words, even if body movements occur, the measures that the user can take vary depending on the type of movement.

[0007] For example, if the body movement is large enough to affect safety, the user can be notified of the movement and immediately provide support to the subject. If the body movement is small enough to affect image quality, artifacts may occur in the image if the user continues imaging. However, if the image can be corrected using the body movement information, the user will not need to re-imaging.

[0008] Patent Document 1 does not describe a method for classifying types of body movements. When it is desired to notify a user of a large body movement that is related to safety, if the movement is detected as a small body movement, the notification will be missed, which may be problematic from the viewpoint of safety. Furthermore, if a small body movement that affects image quality and allows the user to continue the examination is notified as a large body movement, it may actually hinder the technician's workflow.

[0009] Even if the user applies motion correction when a large amount of motion is detected that makes it impossible to correct image artifacts in principle, the process of outputting the corrected image itself becomes a waste, which may hinder the user's workflow.

[0010] As described above, it is important to appropriately classify body movements depending on the application, but at present it is difficult to appropriately classify the body movements of the subject.

[0011] The present invention has been made in view of the above circumstances, and has as its object to provide a nuclear magnetic resonance imaging apparatus and a body movement information processing method that are capable of classifying the body movements of a subject. [Means for solving the problem]

[0012] The invention according to a first aspect is a nuclear magnetic resonance imaging device, which includes an imaging unit that measures nuclear magnetic resonance signals generated by a subject and acquires an image of the subject, and a body movement information processing unit equipped with a processor that processes movement information of a subject placed in the imaging device, the body movement information processing unit including a body movement information calculation unit and a body movement information classification unit, and the processor receives signals from a measurement device that measures the subject's movement information, calculates body movement information from the signals, and classifies the body movement information.

[0013] In the nuclear magnetic resonance imaging apparatus of the second aspect, in the first aspect, the processor classifies the body movement information based on at least two or more types of indices.

[0014] A third aspect of the nuclear magnetic resonance imaging apparatus is the second aspect, wherein the at least two or more indices include the magnitude of body movement and the duration of body movement.

[0015] A fourth aspect of the nuclear magnetic resonance imaging apparatus is the third aspect, wherein the at least two or more indices include the magnitude of body movement, the duration of body movement, and an envelope.

[0016] A fifth aspect of the nuclear magnetic resonance imaging apparatus is the third or fourth aspect, wherein the at least two or more types of indices further include a spatial region where body movement occurs.

[0017] A sixth aspect of the nuclear magnetic resonance imaging apparatus is any one of the first to fifth aspects, wherein the processor selects a partial region of the subject or a characteristic movement of the subject when calculating the body movement information.

[0018] A seventh aspect of the nuclear magnetic resonance imaging apparatus is the same as any of the first to sixth aspects, in which the measuring device is a camera, the signal is an image captured by the camera, and the processor calculates body movement information from changes in the image over time.

[0019] The eighth aspect of the nuclear magnetic resonance imaging apparatus is the same as any of the first to sixth aspects, in which the measuring device is a camera, the signal is an image captured by the camera, and the processor calculates body movement information from the temporal change in the correlation coefficient of the image.

[0020] A ninth aspect of the nuclear magnetic resonance imaging apparatus is the same as any of the first to sixth aspects, in which the measuring device is a stereo camera, the signals are stereo images captured by the stereo camera, and the processor calculates body movement information including three-dimensional information from the stereo images.

[0021] A tenth aspect of the nuclear magnetic resonance imaging apparatus is the nuclear magnetic resonance imaging apparatus of any one of the second to fifth aspects, wherein the processor sets a threshold for each of at least two or more types of indexes and classifies the body movement information based on the threshold.

[0022] An eleventh aspect of the nuclear magnetic resonance imaging apparatus is the tenth aspect, wherein the threshold value is a preset value.

[0023] A twelfth aspect of the nuclear magnetic resonance imaging apparatus is the tenth aspect, wherein the threshold value is a value determined by machine learning using previously collected body movement information as correct answer data.

[0024] A thirteenth aspect of the nuclear magnetic resonance imaging apparatus is the twelfth aspect, wherein the algorithm used for machine learning includes a support vector machine or a decision tree.

[0025] A nuclear magnetic resonance imaging apparatus according to a fourteenth aspect is the nuclear magnetic resonance imaging apparatus according to the twelfth or thirteenth aspect, wherein the supervised data includes extended supervised data generated by data extension.

[0026] A nuclear magnetic resonance imaging apparatus according to a fifteenth aspect is any one of the first to fourteenth aspects, wherein the processor applies different calculation formulas depending on the movement of the subject when calculating the body movement information.

[0027] A nuclear magnetic resonance imaging apparatus according to a sixteenth aspect is the nuclear magnetic resonance imaging apparatus according to any one of the first to fifteenth aspects, wherein the processor outputs a classification result from the body motion information processing section, and / or outputs an alert based on the classification result.

[0028] A seventeenth aspect of the nuclear magnetic resonance imaging apparatus is any one of the first to sixteenth aspects, wherein the processor determines whether or not to apply a body motion correction function based on the classification result from the body motion information processing unit.

[0029] The nuclear magnetic resonance imaging apparatus of an 18th aspect is the nuclear magnetic resonance imaging apparatus of any one of the first to seventeenth aspects, wherein the processor applies a body motion correction function based on the classification result from the body motion information processing unit, and outputs a body motion corrected image based on the body motion information.

[0030] A nuclear magnetic resonance imaging apparatus according to a nineteenth aspect is the nuclear magnetic resonance imaging apparatus according to any one of the first to eighteenth aspects, wherein the processor determines a plurality of regions for measuring motion information of the subject.

[0031] A 20th aspect of the body movement information processing method is a body movement information processing method in which a body movement information processing device equipped with a processor processes movement information of a subject placed on an imaging device, in which the processor receives a signal from a measurement device that measures the subject's movement information, calculates body movement information from the signal, and classifies the body movement information. [Effects of the Invention]

[0032] According to the present invention, it is possible to classify the body movement information of the subject. [Brief explanation of the drawings]

[0033] [Figure 1] FIG. 1 is a diagram showing an overview of a medical imaging device. [Figure 2] FIG. 2 is a diagram showing the relationship between the camera and the MRI imaging device. [Figure 3] FIG. 3 is a diagram showing the flow of processing by the body movement information processing device. [Figure 4]FIG. 4 is a diagram showing an example of a medical imaging apparatus for determining the threshold value of the index. [Figure 5] FIG. 5 is a diagram for explaining the process of calculating body movement information and determining a threshold value. [Figure 6] FIG. 6 is a diagram for explaining a method for extracting feature points from the body movement information extracted in FIG. [Figure 7] FIG. 7 is a graph plotting feature points of body movement information extracted based on FIGS. [Figure 8] FIG. 8 is a diagram for explaining variations in setting the threshold value of the index. [Figure 9] FIG. 9 is a diagram for explaining variations in the method of calculating body movement. [Figure 10] FIG. 10 is a diagram showing an example of body movement calculated based on the motion vector. [Figure 11] FIG. 11 is a diagram for explaining a method for calculating body movement information of a subject. [Figure 12] FIG. 12 is a diagram showing the classification results of the subject's body movement information. [Figure 13] FIG. 13 is a diagram showing an example of a display of the classification results and alerts displayed on the operation unit. [Figure 14] FIG. 14 is a flowchart for explaining the body movement correction process. [Figure 15] FIG. 15 is a diagram for explaining the body movement correction function. [Figure 16] FIG. 16 is a diagram showing a processing flow of another body movement information processing device. [Figure 17] FIG. 17 is a diagram illustrating a mask. [Figure 18] FIG. 18 is a diagram showing the concept of a three-dimensional mask. [Figure 19] FIG. 19 is a diagram for explaining other indexes. DETAILED DESCRIPTION OF THE INVENTION

[0034] Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. In the following description and accompanying drawings, identical components are designated by the same reference numerals, and duplicated explanations will be omitted. Furthermore, when multiple components are listed in the following embodiments, this can be interpreted as including at least one of the multiple components. Hereinafter, embodiments of a nuclear magnetic resonance imaging apparatus and a body movement information processing method of the present invention will be described.

[0035] 1, the medical imaging apparatus 1 includes an MRI apparatus 20 that acquires diagnostic images of a subject, a measurement apparatus 30 that measures motion information of the subject 100, and a body motion information processing apparatus 10 that classifies the body motion information of the subject using the motion information from the measurement apparatus 30. The body motion information processing apparatus 10 is an example of a body motion information processing unit of the present invention. The medical imaging apparatus 1 is an example of a nuclear magnetic resonance imaging apparatus of the present invention.

[0036] The three-dimensional coordinate system shown in FIG. 1 shows an example of the definition of directions in the MRI apparatus 20. The X-axis, Y-axis, and Z-axis of the three-dimensional coordinate system are an example and are not limited to this. For ease of understanding in the following description, the X-axis, Y-axis, and Z-axis in the MRI apparatus 20 are defined in the same direction in all figures. As an example of the three-dimensional coordinate system, the Z-axis direction is the static magnetic field direction, and is the front-back direction of the long axis of the imaging space 140. The Y-axis direction is the up-down direction of the subject. The X-axis direction is the left-right direction of the subject. The MRI apparatus 20 is an example of an imaging unit of the present invention.

[0037] The MRI apparatus 20 is installed in an examination room of an imaging diagnostic facility. In the examination room, the subject 100 is placed on a top plate 144 of a table 142 of a bed device 141.

[0038] The MRI apparatus 20 includes a static magnetic field generating magnet 102, a gradient magnetic field coil 104, and a transmission coil 106. The subject 100 is transported toward the static magnetic field generating magnet 102 of the MRI apparatus 20 by moving a top board 144.

[0039] The static magnetic field generating magnet 102 generates a uniform static magnetic field in an imaging space 140 in which the subject 100 is placed. The gradient magnetic field coil 104 generates a gradient magnetic field in the imaging space 140. The transmission coil 106 generates a high-frequency magnetic field in the imaging space 140 to generate nuclear magnetic resonance signals (NMR (Nuclear Magnetic Resonance) signals) (hereinafter referred to as NMR signals) in the nuclei of atoms constituting the tissue of the subject 100.

[0040] The MRI apparatus 20 includes a high-frequency transmitter 110 , a receiver 114 , a gradient magnetic field power supply 112 , a sequencer 108 , a signal processing unit 116 , and an operation unit 118 .

[0041] The sequencer 108 sends commands to the radio frequency transmitter 110 and the gradient magnetic field power supply 112 in accordance with an imaging sequence (pulse sequence), and appropriately amplified signals are sent to the transmission coil 106 and the gradient magnetic field coil 104, respectively.

[0042] The signal sent to the transmitting coil 106 is applied to the subject 100 as a pulsed radio frequency magnetic field (RF pulse) via the transmitting coil 106. The NMR signal generated from the subject 100 is detected by the coil elements of the receiving coil 150 and is detected by the receiver 114.

[0043] The gradient magnetic field coil 104 is composed of gradient magnetic field coils in three directions, X, Y, and Z, and generates gradient magnetic fields in response to signals from a gradient magnetic field power supply 112 .

[0044] The nuclear magnetic resonance frequency (detection reference frequency f0) used as the detection reference in the receiver 114 is set by the sequencer 108. The sequencer 108 controls each component so that it operates at pre-programmed timing and intensity. Among the programs, a part that particularly describes the timing and intensity of RF pulses, gradient magnetic fields, and signal reception is called a pulse sequence.

[0045] There are various known pulse sequences depending on the purpose, but a detailed description thereof will be omitted here.

[0046] The signal processing unit 116 controls the operation of the MRI apparatus 20 via the sequencer 108, receives the signal detected by the receiver 114, and performs various signal processing such as image reconstruction. The receiver 114 performs quadrature phase detection on the received signal (NMR signal), which is an analog wave, using a set detection reference frequency f0, performs AD (analog-to-digital) conversion, and then transmits the data to the signal processing unit 116. This data is also called the received signal or measurement data.

[0047] The signal processing unit 116 receives various instruction inputs from the operation unit 118 and performs overall control of each unit of the MRI apparatus 20. The signal processing unit 116 also performs processing such as converting the received signals in the spatial frequency domain (k-space) received via the sequencer 108 into an image in real space by inverse Fourier transform, thereby generating an MRI image.

[0048] The signal processing unit 116 is realized by a general-purpose computer such as a personal computer or a microcomputer. The signal processing unit 116 includes a processor (e.g., a CPU (Central Processing Unit)), a ROM (Read Only Memory), a RAM (Random Access Memory), an input / output interface, etc. The processor included in the signal processing unit 116 is an example of the processor of the present invention.

[0049] In the signal processing unit 116, various programs such as a control program stored in the ROM are loaded into the RAM, and the programs loaded into the RAM are executed by the CPU, thereby realizing the functions of each unit of the MRI apparatus 20 and executing various arithmetic processing and control processing via the input / output interface.

[0050] The operation unit 118 includes a mouse, a keyboard, a display device, a gantry monitor, etc. The operation unit 118 functions as part of a GUI (Graphical User Interface) that accepts user input. The display device of the operation unit 118 displays MRI images and the like generated by the signal processing unit 116.

[0051] The MRI apparatus 20 is an example of an imaging apparatus of the present invention. Although the MRI apparatus 20 is exemplified as an imaging apparatus, other imaging apparatuses such as a CT (Computed Tomography) apparatus, an X-ray diagnostic apparatus, and a PET (Positron Emission Tomography) apparatus may be used. The type of imaging apparatus is not particularly limited as long as it has an imaging space and can generate an image of a subject.

[0052] The measuring device 30 is a device that acquires biometric information of the subject 100 in a non-contact manner during an examination using the MRI device 20. The measuring device 30 is composed of, for example, a camera. When the measuring device 30 is a camera, moving image data (time-series image data) captured by the camera over a predetermined range of the subject 100 is an example of a signal of movement information measured by the measuring device 30. The moving image data is an example of an image captured by a camera of the present invention.

[0053] FIG. 2 shows an example in which two cameras 30A and 30B are installed as the measurement device 30 in the imaging space 140 of the MRI device 20. A top plate 144 on which the subject 100 is placed is inserted into the imaging space 140 of the MRI device 20. The camera 30A is installed upward near the insertion end of the imaging space 140. The camera 30B is installed upward on the opposite side of the imaging space 140 from the insertion end. The cameras 30A and 30B are installed at positions that capture the subject 100 from diagonally above. As shown in FIG. 2, by attaching the cameras 30A and 30B to both sides of the long axis of the imaging space 140, it is possible to select and use moving image data suitable for detecting movement information of the subject 100. For example, in the example shown in FIG. 2, the camera 30A at the insertion end is close to the abdomen of the subject 100, and the camera 30B on the opposite side is close to the face of the subject 100. The cameras 30A and 30B can easily acquire signals of their respective movement information.

[0054] The following description will be given taking the example of cameras 30A and 30B being monocular cameras. When cameras 30A and 30B are monocular cameras, the subject's body movements are calculated as relative values. Note that measurement device 30 may also be a stereo camera. In the case of a stereo camera, three-dimensional information can be obtained from stereo images. Therefore, the distance to the camera can be obtained, and the body movement waveform can be obtained as an absolute value.

[0055] Returning to Fig. 1, the body movement information processing device 10 is a device that receives signals of movement information of the subject 100 from the measurement device 30, calculates body movement information from the received signals, and classifies the calculated body movement information. The body movement information processing device 10 can present the classification results of the body movement information to the user and can also output them to the signal processing unit 116 of the MRI device 20. By knowing the classification results of the body movement information, the user can determine whether to continue imaging, whether to correct for body movement using the body movement information, etc.

[0056] The body movement information processing device 10 may classify the body movement information using at least two or more indices. The body movement information processing device 10 may set thresholds for the indices in advance and classify the body movement information. When calculating body movement information from a received signal, the body movement information processing device 10 may determine a spatial region for calculation.

[0057] To perform these functions, the body movement information processing device 10 of the embodiment includes an index setting unit 11, a body movement information calculation unit 13, a body movement information classification unit 15, etc. The body movement information processing device 10 can be configured, for example, by a computer equipped with a memory and a processor (CPU: Central Processing Unit) and / or a GPU (Graphic Processing Unit). The functions of each unit of the body movement information processing device 10 can be realized by the CPU or the like loading a program that realizes the function. Furthermore, some of the functions may be realized by a programmable IC such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). These processors included in the body movement information processing device 10 are examples of the processor of the present invention.

[0058] 1 shows the case where the body movement information processing device 10 is a device separate from the MRI device 20, but the function of the body movement information processing device 10 can also be incorporated into the signal processing unit 116 of the MRI device 20, and such a configuration is also included in the medical imaging device of the present invention. Therefore, the signal processing unit 116 can have the function of the processor of the body movement information processing device 10.

[0059] Hereinafter, an embodiment of the processing of the body movement information processing device 10 will be described using as an example the case where the imaging device is an MRI device 20. Fig. 3 is a diagram showing the flow of processing of the body movement information processing device.

[0060] First, an index for classifying body movement information is determined (step S1). Specifically, an index to be applied to classifying body movement information is set by the index setting unit 11. Furthermore, the index setting unit 11 can set a threshold value for each set index.

[0061] Here, the movement information of the subject 100 is information obtained from the actual movement of the subject 100. The signal is an electrical signal output from the measurement device 30 that measures the movement information of the subject 100, and includes moving image data, etc. The body movement waveform (sometimes simply referred to as body movement) is information calculated by a processor provided in the body movement information processing device 10 based on the signal output from the measurement device 30, and corresponds to the movement information. The body movement information is information that enables body movement to be classified, and includes at least two evaluation values ​​corresponding to indices. The body movement information can be calculated based on the body movement waveform.

[0062] Next, the processing in the index setting unit 11 will be described. The movement information of the subject 100 is calculated as body movement information as will be described later. When classifying the body movement information based on an index, it is desirable that the index be perceived from the movement information of the subject 100. The perceived movement information of the subject 100 can be roughly divided into the magnitude of the movement, the duration of the movement, and the spatial region where the movement occurs, and it is desirable to set the magnitude of the body movement, the duration of the body movement, and the spatial region where the body movement occurs as indexes for classifying the body movement information.

[0063] By applying at least two or more of these indices to classify the body movement information, it becomes possible to determine not only the presence or absence of movement of the subject 100 but also the type of movement of the subject 100. For example, while imaging the subject 100, the user can know whether the movement of the subject 100 is a large movement that affects safety or a small movement that affects the quality of the captured image.

[0064] Next, we will explain how to determine the thresholds for two types of indices, using the magnitude and duration of the subject's body movement as examples. The determined thresholds for each indices are used to classify body movement information, which will be described later.

[0065] Fig. 4 is a diagram showing an example of a medical imaging apparatus 1 for determining threshold values ​​of indices, and Fig. 5 is a diagram for explaining calculation of body movement information and setting of threshold values.

[0066] The MRI apparatus 20, cameras 30A and 30B, and body movement information processing device 10 shown in Fig. 4 are basically the same as those in Figs. 1 and 2. The body movement information processing device 10 is connected to an operation unit 12 including a display device 12A and an input device 12B. Note that the operation unit 12 can also be used as the operation unit 118 of the MRI apparatus 20. The MRI apparatus 20 for examining the subject 100 and the like are used to determine the threshold value. Applying this threshold value ensures the accuracy of the classification of body movement information, which will be described later.

[0067] The subject 100A is a subject model for collecting body movement information of a sample by the body movement information processing device 10. The subject 100A reproduces movements that may occur in the subject 100 according to instructions from the user U. The subject 100A reproduces, for example, movements that involve large changes and last for a short time, such as coughing or sneezing, as well as movements that involve small changes and last for a certain period of time, such as body squirming. The subject 100A can also reproduce movements that involve large changes and last for a certain period of time, such as movements that may be related to safety, as well as movements that involve small changes and last for a short period of time that do not affect imaging.

[0068] First, upon receiving signals from cameras 30A and 30B, body movement information calculation unit 13 calculates motion vectors using known computer vision techniques such as optical flow, and then calculates body movement information. The calculation of body movement information by body movement information calculation unit 13 will be described later. Here, a method for setting indices and thresholds based on the calculated body movement information will be described.

[0069] 5-1 in Figure 5 shows a body movement waveform (also called body movement) obtained from the motion vector results, and is a graph plotting the calculated results with time (s) on the horizontal axis and magnitude (amplitude) on the vertical axis. The body movement waveform in 5-1 serves as the basis for determining the threshold value of the index. The body movement information calculation unit 13 calculates the body movement waveform and body movement information based on an index suitable for classification by the body movement information classification unit 15.

[0070] To calculate body movement information from the body movement waveform in 5-1, for example, waveform data including large and small amplitude body movements over 10 seconds (101 points) is used. Next, as shown in the body movement waveform in 5-2 of Fig. 5, frames (e.g., frame W1, frame W2, etc.) determined by the amplitude (10,000 points from 0.01 to 100) and duration (80 points from 0.1 to 8 seconds) are set for the waveform, and the body movement information is extracted. For example, a graph of the waveform shown in Fig. 5 is displayed on the display device 12A of the operation unit 12 as shown in Fig. 4.

[0071] Different amplitude P1 and duration P2 are set for frames W1 and W2. For example, if a body movement waveform exceeds a set amplitude, it is determined that body movement has occurred, and if it exceeds the set amplitude within a set duration, it is determined that body movement is continuing. If a body movement waveform does not exceed the set amplitude within a set duration, it is determined that body movement has ended. If the conditions of frame W1 are met, it is recognized as body movement for one waveform, and body movement information is extracted. Similarly, if the conditions of frame W2 are met, it is recognized as body movement for another waveform, and body movement information is extracted. The body movement information has amplitude and duration information corresponding to the index.

[0072] When determining the duration, the amplitude of the body movement waveform may fluctuate, repeatedly exceeding and not exceeding the set amplitude, making it difficult to determine the duration. In such cases, a time interval P3 for determining duration is set. For example, consider a case where the waveform has two peaks. The time from when the body movement waveform exceeds the set amplitude P1, falls below it, and then exceeds the set amplitude P1 again is calculated. If this time is shorter than the time interval P3 for determining duration, the two peaks are determined to be one continuous body movement. However, if it is greater than P3, the two peaks are determined to be separate body movement waveforms rather than a continuous body movement.

[0073] FIG. 6 is a diagram illustrating a method for extracting feature points of body movement information from the body movement information (body movement for one waveform) extracted in FIG. 5. 6-1 in FIG. 6 shows the body movement information extracted in FIG. 5 mapped onto a graph with magnitude (amplitude) on the vertical axis and time on the horizontal axis. When mapped as shown in 6-1 in FIG. 6, the body movement information can be represented by region AR1. As a result, the graph is divided into region AR1 and region AR2 outside region AR1. That is, region AR1 corresponds to cases where the set amplitude or duration value is exceeded, and region AR2 corresponds to cases where the set amplitude or duration value is not exceeded. The boundary between region AR1 and region AR2 provides information indicating the characteristics of the amplitude or duration of a certain body movement waveform.

[0074] 6-2 in Figure 6 is a graph plotting boundary points on the boundary between area AR1 and area AR2. Each boundary point can be expressed as a value of the product of magnitude (amplitude) and time.

[0075] 6-3 in Figure 6 is a graph showing only the boundary point with the largest product of magnitude (amplitude) and time among all the boundary points. 6-4 in Figure 6 is an enlarged view of 6-3 in Figure 6. By extracting the boundary point with the largest product as a feature point for body movement information, the body movement information can be defined as a single feature point with magnitude (amplitude) and time. This makes it possible to exclude cases with extremely small magnitude (amplitude) or time (cases that may contain a lot of noise). Here, the boundary point with the largest product is defined as the feature point for body movement information, but feature points for body movement information are not limited to this and can be selected appropriately by the user.

[0076] 7 plots the feature points of the plurality of pieces of body movement information extracted based on FIGS. 5 and 6 on a graph (also called a classification graph) with the vertical axis representing magnitude (amplitude) and the horizontal axis representing duration (s). The body movement information extracted by body movement information calculation unit 13 is input to index setting unit 11. Index setting unit 11 creates a classification graph shown in 7-1 of FIG. 7 based on the feature points of the plurality of pieces of body movement information, for example.

[0077] Here, the ● marks indicate feature points (body movement information) corresponding to movements of the subject 100A that are large in amplitude and long in duration. The × marks indicate feature points (body movement information) corresponding to movements of the subject 100A that are large in amplitude and short in duration, movements of the subject 100A that are small in amplitude and short in duration, and movements of the subject 100A that are small in amplitude and long in duration. The body movement information of large movements of the subject 100A (● marks) includes, for example, movements of the subject that are in a range that affects safety and makes it difficult to continue the examination. For example, there is the body movement shown in the body movement waveform 7-2 in Figure 7, which has a large magnitude (amplitude) and a long duration.

[0078] The body movement information of the subject 100A with small movements (marked with an x) includes subject movements that allow the examination to continue and that are correctable but affect the image quality of the captured image, or subject movements that allow the examination to continue and do not affect the image quality of the captured image. For example, the body movement shown in the body movement waveform of 7-3 in Fig. 7 has a small magnitude (amplitude) and a short duration.

[0079] In 7-1 of FIG. 7, only the feature points extracted in 6-3 of FIG. 6 are plotted, but it is also possible to plot all of the boundary points shown in 6-2 of FIG.

[0080] Next, variations in setting the threshold value of the index will be described with reference to Fig. 8. The index threshold value is set by the index setting unit 11.

[0081] The classification graph 8-1 in Figure 8 is a diagram showing a first variation of the threshold setting method, the classification graph 8-2 ​​in Figure 8 is a diagram showing a second variation of the threshold setting method, and the classification graph 8-3 in Figure 8 is a diagram showing a third variation of the threshold setting method.

[0082] The first variation of the threshold setting method is a rule-based method. As shown in 8-1, body movement information can be divided into a first group of body movement information indicated by a dotted line with a ● mark, and a second group of body movement information indicated by a dotted line with an × mark. It can be understood that the first and second groups are classified by setting thresholds Th1 and Th2 for magnitude (amplitude) and time (duration), respectively. That is, threshold Th1 is determined for the magnitude (amplitude) index, and threshold Th2 is determined for the time (duration) index. In the first variation, the user determines the thresholds themselves, making the classification process clear.

[0083] Next, a method for setting a threshold using a machine learning technique will be described. In the machine learning technique, a threshold is set using extracted body movement information as correct data.

[0084] The second variation of the threshold setting method applies a decision tree, a machine learning technique. A decision tree is a machine learning technique that uses a tree structure to perform classification and other tasks. Its purpose is to find optimal conditions for dividing given data. A decision tree has a tree structure, with leaves representing classifications and branches representing feature quantities leading to the leaves. Given correct answer data, the decision tree algorithm finds optimal conditions for dividing the data. In the example of 8-2 in Figure 8, the decision tree discovers a condition where y1 is the magnitude (amplitude) and x1 is the duration, and y1 is determined as the threshold Th1 and x1 as the threshold Th2. The decision tree method can improve classification accuracy and clarify the classification process compared to the first variation of the threshold setting method. Since decision trees are a well-known technique, detailed explanations are omitted.

[0085] The third variation of the threshold setting method applies a support vector machine (SVM), a machine learning technique. The purpose of SVM is to find a boundary (hyperplane) that optimally classifies given data. SVM determines a division line by defining support vectors. SVM can process nonlinear data using margin maximization and kernel methods. In Figure 8-3, lines are drawn near the support vectors marked with a circle and the support vector marked with a circle. The two lines separate the body movement information marked with a circle and the body movement information marked with a cross. A dashed line L1 is drawn between the two lines. The dashed line L determines the thresholds for the two indicators, allowing the body movement information to be classified. A stepped line can also be drawn instead of the line L1. Compared to the first and second variations of the threshold setting method, SVM can improve classification accuracy and clarify the classification process. Since SVM is a well-known technology, a detailed description is omitted.

[0086] In the second and third variations of the threshold setting method, if the correct answer data does not contain sufficient body movement information, extended correct answer data for the body movement information may be generated by data expansion. For example, the data may be expanded in the direction of the body movement magnitude within the allowable range, or in the direction of the body movement duration within the allowable range. In this case, for example, data may be generated in which the body movement magnitude is 0.5 times, 0.6 times, . . . 1.5 times larger, or data in which the duration is 2 times or 3 times larger. When the correct answer data includes the extended correct answer data for the body movement information, the classification accuracy is improved. Note that data expansion is a well-known technique, and therefore a description thereof will be omitted.

[0087] The thresholds determined in the first to third variations are stored in a memory or the like provided in the body movement information processing apparatus 10, completing the setting of the thresholds. The thresholds are used to classify the body movement information.

[0088] In the embodiment, a case has been described in which the threshold value is determined by a machine learning function implemented in the index setting unit 11, but this is not limiting. For example, the machine learning function may be implemented in another system, and the threshold value may be calculated when body movement information is input. That is, the index setting unit 11 may be capable of selecting an index and setting a threshold value in response to a user instruction or automatically. Furthermore, machine learning may be performed in advance before product shipment, and the threshold value calculated by machine learning may be stored in the index setting unit 11.

[0089] After the indices are set (step S1), the subject 100 is imaged using the medical imaging device 1. Once the indices are set, the subject 100 can be examined in subsequent examinations without setting the indices (step S1).

[0090] 3, after placing the indices (step S1), a motion information signal is received (step S2). Specifically, while an MRI image of the subject 100 is being captured, the body motion information calculation unit 13 receives a signal of moving image data of the subject 100 captured by the cameras 30A and 30B.

[0091] After receiving the video data signal (step S2), body movement information is calculated (step S3). Specifically, body movement information calculation unit 13 calculates the body movement from the motion vector and the velocity vector.

[0092] After receiving the video data signal, the body movement information calculation unit 13 calculates a motion vector. The motion vector can be calculated using a known computer vision technique such as optical flow. Optical flow calculates a motion vector for each pixel of an image based on temporal changes in the video data. Known optical flow algorithms include the Ferneback method and the Lucas-Kanade method, and either method may be used.

[0093] The calculated result is a velocity vector that represents the change in pixel position between frames for each pixel. For example, as shown in Figure 9, when the length and width of the image IM are the x and y directions, the velocity is calculated as the velocity in the x direction (Vx) and the velocity in the y direction (Vy).

[0094] Using these velocity vectors Vx and Vy, the body movement is calculated using the following equation. In this case, when the subject 100 is lying on a table, the respiratory movement of the subject 100 is mainly perpendicular to the horizontal plane (table surface). On the other hand, the movements of the subject 100 other than the respiratory movement are irregular (or random) movements, which is a difference in the characteristics of the two movements. Calculations can be performed using different formulas depending on the type of body movement. For example, the type of body movement can be determined based on information about the examination site when the user sets the examination site at the start of the examination.

[0095] For sudden movements, Vmotion can be calculated using, for example, the following formula (1) or formula (2).

[0096]

number

[0097]

number

[0098] When it is desired to observe body movement (respiratory movement) in the direction of chest movement with high sensitivity, Vabd_motion can be calculated by the following equation (3), or Vmotion can be calculated by equation (4) using equation (3).

[0099]

number

[0100]

number

[0101] According to equation (4), integrating in the time direction for the number of data points in tcount has the effect of improving sensitivity. θ in the equation refers to the angle between the horizontal direction and the direction of respiratory movement (also called the abdomen-back direction, or the A (anterior)-P (posterior) direction). In this example, horizontal and vertical velocity components are calculated using optical flow, and θ is set based on this information to increase the sensitivity of respiratory movement detection.

[0102] Note that θ in the formula is the angle of the XY plane with respect to the horizontal plane, and varies depending on the positional relationship with the cameras 30A and 30B, the mounting angle of the cameras, etc. θ may also be determined after analyzing the direction of respiratory movement for each subject. θ can be set appropriately so that respiratory movement, other movements, and body movements can be detected with high sensitivity, and different values ​​may be used for each position (pixel). The formula for expressing body movement may be any that expresses the body movement to be detected, and is not limited to (1) to (4).

[0103] The body movement information calculation unit 13 can resize the images of the video data as needed to downsize the original image size. For example, the original size of the camera image can be reduced to about 1 / 2 to 1 / 10. By reducing the image size of the video data in this way, the computational load of subsequent processing such as motion vector calculation can be reduced, enabling faster computation (real-time presentation). Furthermore, in order to divide one image into multiple regions of the same pixel size in subsequent processing, the resized image may be cut off at the edges of the image, making it divisible by the number of divisions both vertically and horizontally.

[0104] The resized image is divided into small regions, and for each divided small region, the average values ​​of the velocity vectors and motion vectors found for each pixel in the small region are calculated. For example, if the pixel size of the resized camera image after fractional parts is 720 pixels wide x 540 pixels high, and this is divided into 60 regions, each of which has a size of 72 x 90 pixels, the average values ​​of the velocity vectors and motion vectors are calculated for each small region of this size.

[0105] It should be noted that body movement information can be calculated from temporal changes in the correlation coefficient of images captured by a camera, without being limited to optical flow. For example, first, target image data is acquired. Next, the correlation coefficient is calculated between successive image data (frames). Next, the correlation coefficient can be calculated pixel by pixel. The relationship between corresponding pixels is evaluated. The temporal changes in the calculated correlation coefficient are analyzed to detect changes. Body movement waveforms can be calculated based on these changes in the correlation coefficient.

[0106] 10 is a diagram showing an example of body movement and body movement information calculated by body movement information calculation unit 13 based on the motion vector. Each body movement waveform in FIG. 10 is a graph calculated for each small region, with the vertical axis representing magnitude (amplitude) and the horizontal axis representing time. Each body movement waveform is calculated by body movement information calculation unit 13 based on moving image data captured by cameras 30A and 30B of the entire subject 100. In this example, body movements for 60 small regions are calculated.

[0107] Next, a method for calculating body movement information according to an index will be described with reference to Fig. 11. Fig. 11-1 shows an example of calculating body movement information using two indexes (amplitude and duration), and Fig. 11-2 shows an example of calculating body movement information using three indexes (amplitude, time, and space).

[0108] 11-1 in FIG. 11 is a body movement waveform calculated by the body movement information calculation unit 13 for the body movement of a specified region of the subject 100. In other words, a partial region of the subject 100 is the monitoring target. The body movement information calculation unit 13 calculates body movement information from the amplitude and duration, which are indicators. The calculated body movement information is classified by the body movement information classification unit 15. The body movement information calculation unit 13 calculates body movement information by treating the body movement waveform surrounded by a frame W as one waveform.

[0109] In 11-2 of FIG. 11, the body movement information calculation unit 13 simultaneously calculates body movement waveforms for multiple regions (spaces) 1, 2, ..., N of the subject 100. That is, in 11-2, the body movement waveform is calculated using space as one of the indices. When simultaneously calculating body movement waveforms for multiple regions, it is important to determine from which region's body movement waveform to calculate body movement information. As one determination method, in 11-2, at each time, the maximum value (maximum amplitude) of all regions is selected and the body movement waveform is calculated. The region max is a body movement waveform obtained by plotting the maximum amplitude over time. In other words, the region max is a body movement waveform that selects the characteristic movement of the subject 100. Therefore, the region max is a composite body movement waveform that includes the movements of multiple regions of the subject 100. As another determination method, one region with large body movement may be selected based on the body movement waveform over a certain period of time.

[0110] Next, similarly to 11-1, body movement information calculation unit 13 calculates body movement information from area max using the amplitude and duration as indices. Body movement information calculation unit 13 calculates body movement information by treating the body movement waveform surrounded by frame W as one waveform. The calculated body movement information is classified by body movement information classification unit 15.

[0111] As another method of determination, one area may be selected as a representative area from among a plurality of areas (spaces) 1, 2, . . . N. The user can appropriately set the area to be monitored depending on the part to be imaged.

[0112] After calculating the body movement information (step S3), the body movement information is classified (step S4). Specifically, the body movement information classifier 15 classifies the body movement information (including information on amplitude and duration) based on indices. FIG. 12 shows the classification results of the body movement information classified by the body movement information classifier 15. In 12-1 of FIG. 12, the body movement information classified regardless of the time of occurrence of the body movement is plotted on a graph with the vertical axis representing magnitude (amplitude) and the horizontal axis representing time. Furthermore, the thresholds Th1 and Th2 set by the index setting unit 11 are also displayed. The body movement information (●) in the upper right region A surrounded by the thresholds Th1 and Th2 (amplitude equal to or greater than the threshold Th1 and duration equal to or greater than the threshold Th2) corresponds to large movements of the subject 100, as shown in 12-2. Furthermore, the body movement information (×) in region B outside region A (amplitude less than the threshold Th1 or duration less than the threshold Th2) corresponds to small movements of the subject 100, as shown in 12-3.

[0113] According to the classification results, it can be seen that the body movement information of the subject 100 was classified by the magnitude (amplitude) of the body movement of the threshold value Th1 and the duration of the body movement of the threshold value Th2.

[0114] Next, after classifying the body movement information (step S4), the classification result is transferred (step S5). Specifically, the body movement information processing device 10 transfers the calculated body movement waveform and the classification result of the body movement information to the signal processing unit 116 of the MRI device 20. The body movement waveform includes the time when the body movement occurred and the magnitude of the amplitude.

[0115] Next, after transferring the classification result (step S5), the classification result and / or an alert are output (step S6). Specifically, the signal processing unit 116 outputs the calculated body movement and the classification result of the body movement information to a display device (not shown) of the operation unit 118. The signal processing unit 116 can output this information to a display device, a printing device, or a storage device.

[0116] 13 is a diagram showing an example of a body movement waveform 130, a classification result 131, and an alert 132 displayed on the display device 118A of the operation unit 118. As shown in FIG. 13, an image 133 captured by the MRI device 20 is displayed on the display device 118A. In addition, the body movement waveform 130 and the classification result 131 are displayed. The body movement waveform 130 and the classification result 131 are associated with the time of occurrence. Therefore, the user can know when the body movement information on the classification result 131 is caused by the body movement that occurred.

[0117] The method of outputting the alert 132 is not particularly limited, but in this example, the alert 132 is displayed below the image 133 as an icon and text, for example. The alert 132 may be either an icon or text. The alert 132 may also change the icon and text display (size, color, text content, etc.) depending on the classification. A warning sound may also be emitted, or a voice alert may be used. These may also be combined as appropriate.

[0118] By outputting an alert 132 according to the classification, it is possible to notify a user who is concentrating on other tasks that an abnormal movement has occurred. Also, when the subject sneezes or coughs temporarily according to the classification, imaging can be continued from the viewpoint of the subject's safety, and unnecessary interruptions to imaging can be avoided. On the other hand, when imaging should be interrupted from the viewpoint of the subject's safety, the user can avoid danger such as the subject falling.

[0119] Next, the use of the classification results of the body motion waveform and body motion information, other than alerts, will be described with reference to FIGS. 14 and 15, with reference to body motion correction. FIG. 14 is a flowchart for explaining the body motion correction process. FIG. 15 is a diagram for explaining the body motion correction function. Explanations of parts common to the processing flows already described may be omitted. Here, body motion correction refers to reconstructing an image with reduced artifacts from the generated MRI image.

[0120] 14, a motion information signal is received (step S11), then body motion information is calculated (step S12), and then the body motion information is classified (step S13). Specifically, the same processes as in steps S1, S2, and S3 are executed.

[0121] The classification result is transferred to the signal processing unit 116, which determines whether or not the body movement can be corrected (step S14). The signal processing unit 116 can determine whether or not body movement correction is possible based on the classification result. In making this determination, a threshold value set in advance for the index is used. The signal processing unit 116 determines whether or not body movement correction is possible by comprehensively determining, for example, the amplitude and duration, which are threshold values.

[0122] If the amplitude is small and not executing the correction function does not affect the image quality, the correction function is not executed. Also, if the amplitude is large and executing the correction function does not improve the image quality, the correction function is not executed. The signal processing unit 116 can determine whether or not to execute body motion correction, without being limited to these cases.

[0123] If it is determined that motion correction is not possible (No in step S14), the process ends without executing motion correction. After the process ends, an MRI image without motion correction is reconstructed, or an MRI image is not reconstructed.

[0124] On the other hand, if it is determined that body movement correction is possible (Yes in step S14), body movement correction processing is executed.

[0125] As shown in 15-1 of Figure 15, in the MRI device 20, an MRI image of a subject is generated by filling measurement lines (indicated by solid arrows) in the measurement data space (k space) according to a certain measurement order (indicated by dashed arrows).

[0126] As shown in 15-2 of FIG. 15, an MRI image 1501 for which it is determined that motion correction is possible is input to the signal processing unit .

[0127] Next, a body motion removal mask is created (step S15). Specifically, as shown in 15-2 of FIG. 15, a body motion removal mask 1503 is created by the signal processing unit 116 based on a body motion waveform 1502 calculated by the camera. The body motion waveform 1502 indicates the body motion waveform while the MRI image is being captured. The body motion removal mask 1503 extracts body motion that causes artifacts. The body motion removal mask 1503 is created taking into account the amplitude and duration of the body motion, etc.

[0128] Next, signals are removed from the k-space (step S16). Specifically, as shown in 15-2 in Fig. 15, the signals are converted into k-space information, and k-space data 1504 is generated from which measurement lines collected at the time when the body movement waveform occurred are removed.

[0129] Next, the removed signal is interpolated (step S17). For example, the removed signal is interpolated by applying a known interpolation method such as iterative reconstruction, which performs repeated calculations while maintaining the consistency of the measurement data.

[0130] Finally, the motion-corrected image is output (step S18). As shown in 15-2 of FIG. 15, an artifact-reduced MRI image 1505 is reconstructed and output. As a result, the image quality of the MRI image can be improved. When creating a motion removal mask, referring to the classification results of the motion information enables more accurate motion correction. The artifact-reduced MRI image 1505 is an example of a motion-corrected image of the present invention.

[0131] Next, a case where body motion information for each different ROI (Region of Interest) is used for motion information of the subject 100 using the medical imaging device 1 will be described with reference to Fig. 16. Explanation of parts common to the processing flow already described may be omitted.

[0132] 16, moving images captured by cameras 30A and 30B are input to body movement information processing apparatus 10 (step S21). In this example, an index and a threshold value have already been set.

[0133] Next, the body movement information calculation unit 13 calculates the optical flow based on the signal of the moving image (step S22), and calculates the motion vector (step S23).The average value of the velocity vector of each small region of the image is calculated from the motion vector (step S24).

[0134] Next, the body movement waveform and body movement information are calculated based on the maximum value within ROI1 (step S25), and the body movement waveform and body movement information are calculated based on the maximum value within ROI2 (step S28). Steps S25 and S28 can be executed in parallel. ROI1 is a region for monitoring large body movements related to safety, such as the entire imaging section or the entire subject. Meanwhile, ROI2 is a region for monitoring small body movements related to image quality, such as the imaging region of the MRI device or the examined region of the subject (head, abdomen, joints, etc.).

[0135] Next, a threshold value of body movement information (subject movement) that may cause the examination to be interrupted is determined (step S26). The classification result by the body movement information classification unit 15 can be used. If the determination result shows that the threshold value has been exceeded, the classification result and an alert are output (step S27). The user can check this output to ensure the safety of the subject.

[0136] In parallel with step S25, a threshold for body movement information around the examination region is determined (step S29). As in step S26, the classification result by the body movement information classification unit 15 can be used. If the determination result shows that the threshold is exceeded, the classification result and an alert are output (step S30). Since ROI2 is the region in which the subject is being imaged, a body movement waveform is output regardless of the determination result (step S31). Steps S25 and S29 are performed in parallel, but step S25 monitors a region related to safety (ROI1). Therefore, the output of the alert in step S27 is executed with priority over the alert in step S30.

[0137] Although the case where a body movement waveform is calculated for each ROI has been described, a mask can be used to extract the body movement information in order to classify it. The body movement information processing apparatus 10 can generate a mask in advance according to a user's instruction. A plurality of regions for measuring the movement information of the subject 100 can be determined. The user can cause the body movement information processing apparatus 10 to generate a mask in advance via the signal processing unit 116 from the operation unit 118 of the MRI apparatus 20. Even in this case, a plurality of regions for measuring the movement information of the subject 100 can be determined.

[0138] Examples of masks for extracting two regions are shown in Fig. 17. Fig. 17 shows an example in which two types of masks 122 and 123 are applied to a body movement region map 121 calculated for each small region of a camera image (after resizing) 120. Of these, mask 122 extracts the imaging region and corresponds to ROI2. In the case of the MRI apparatus 20, a mask can be used in which the region of the subject 100 wearing the receive coil 150 is set to 1 and the rest is set to 0. The position where the receive coil 150 is worn may be determined from the images of cameras 30A and 30B, or may be determined based on the center position of the imaging space and the size of the receive coil 150, since in the MRI apparatus 20 the subject 100 is positioned so that the center position of the receive coil 150 is approximately near the center of the imaging space.

[0139] On the other hand, the mask 123 extracts a wider area including the imaging area and corresponds to ROI 1. It can be a mask in which the area where the subject 100 is present is set to 1 and the rest is set to 0.

[0140] By monitoring the body movement in the region extracted by the mask 122, it is possible to detect body movement that directly affects imaging, and to feed back this information to the imaging. Also, by monitoring the body movement in the region extracted by the mask 123, it is possible to detect any significant movement of the subject 100 during the examination, even if it does not directly affect imaging.

[0141] However, the mask to be generated is not limited to the above example, and one or more (two or more) regions may be set in consideration of the examination site and the mobility of the subject 100 (child, etc.). For example, in the case of a receiving coil in which multiple small coils are connected, the region in which the receiving coil is attached may be divided into multiple regions, or may be divided into a region on the head side including the receiving coil and a region on the leg side including the receiving coil.

[0142] Furthermore, the mask need not be a binary mask of 1 / 0, but may be one with a predetermined weighting that takes into account the positional relationship with the cameras, the sensitivity distribution of the receiving coils, etc. For example, the farther the distance from cameras 30A and 30B, the smaller the body movement that is detected, so the weighting may be increased and the closer the distance, the smaller the weighting. Also, since the sensitivity of receiving coil 150 is generally high in the center and the influence of body movement in that area is considered to be greater, the weighting may be increased in the center and decreased in the periphery.

[0143] Although the above description has been given of an example of a spatial mask, a temporal element and the magnitude of body movement may also be added. The temporal element is, for example, a temporal mask that indicates whether the mask is being taken at the time of imaging or during an examination including before and after imaging.

[0144] FIG. 18 shows the concept of a three-dimensional mask 124 that combines a time element and the magnitude of body movement. Using such a three-dimensional mask allows for monitoring of an appropriate region. Specifically, a three-dimensional mask can be used that combines a mask that selects the sensitivity distribution region of the receiving coil as the spatial mask, a mask that selects the imaging time as the temporal mask, and a mask that selects a first range as the amplitude magnitude mask. This allows for use as a mask corresponding to ROI2. Alternatively, a three-dimensional mask can be used that combines a mask that selects the entire imaging space as the spatial mask, a mask that selects the time during examination as the temporal mask, and a mask that selects a second range equal to or greater than the first range as the amplitude magnitude mask. This allows for use as a mask corresponding to ROI1.

[0145] In this example, body movement information was calculated from the body movement waveform using the magnitude (amplitude) and duration as indices. A case in which an envelope is used as another piece of information will be described with reference to Fig. 19. 19-1 in Fig. 19 shows a body movement waveform 200, an envelope 201, and a straight line 202. 19-2 in Fig. 19 shows a body movement waveform 203, an envelope 204, and a straight line 205 that are different from those in 19-1.

[0146] Body movement waveform 200 shows a waveform with a small rising edge. Envelope 201 is a line that surrounds the outside of body movement waveform 200 and shows a gentle mountain-like shape. Straight line 202 is tangent to envelope 201.

[0147] On the other hand, body movement waveform 203 shows a waveform with a large rising edge. Envelope 204 is a line that surrounds the outside of body movement waveform 203 and shows a shape with a steep slope. Straight line 205 is tangent to envelope 204.

[0148] The slopes of the lines 202 and 205 are calculated from the envelopes 201 and 204, and the values ​​can be used to classify the body movement information.

[0149] Furthermore, it goes without saying that the present invention is not limited to the above-described embodiment, and various modifications are possible. [Explanation of symbols]

[0150] 1 Medical imaging device 10 Body movement information processing device 11 Index setting section 13 Body movement information calculation unit 15 Body movement information classification unit 20 MRI machine 30 Measuring Equipment 30A, 30B Camera 100, 100A specimen

Claims

1. an imaging unit that measures a nuclear magnetic resonance signal generated by a subject and acquires an image of the subject; a body motion information processing unit including a processor that processes motion information of the subject placed in the imaging device; The body movement information processing unit A body movement information calculation unit and a body movement information classification unit are provided, The processor: receiving a signal from a measurement device that measures motion information of the subject, and calculating body motion information from the signal; A nuclear magnetic resonance imaging device that classifies the body movement information.

2. The processor: The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the body movement information is classified based on at least two or more types of indices.

3. The nuclear magnetic resonance imaging apparatus according to claim 2 , wherein the at least two or more indices include a magnitude of body movement and a duration of body movement.

4. The nuclear magnetic resonance imaging apparatus according to claim 3 , wherein the at least two or more indices include a magnitude of the body movement, a duration of the body movement, and an envelope of the body movement.

5. 5. The nuclear magnetic resonance imaging apparatus according to claim 3, wherein the at least two or more types of indices further include a spatial region where a body movement occurs.

6. The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor selects a partial region of the subject or a characteristic movement of the subject when calculating the body movement information.

7. the measurement device is a camera, and the signal is an image captured by the camera, The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor calculates the body movement information from a temporal change in the image.

8. the measurement device is a camera, and the signal is an image captured by the camera, The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor calculates the body motion information from a temporal change in a correlation coefficient of the images.

9. the measurement device is a stereo camera, the signal is a stereo image captured by the stereo camera, The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor calculates the body motion information including three-dimensional information from the stereo images.

10. The nuclear magnetic resonance imaging apparatus according to claim 2 , wherein the processor sets a threshold value for each of the at least two or more types of indexes, and classifies the body motion information based on the threshold value.

11. The nuclear magnetic resonance imaging apparatus according to claim 10 , wherein the threshold value is a preset value.

12. The nuclear magnetic resonance imaging apparatus according to claim 10 , wherein the threshold value is a value determined by machine learning using previously collected body movement information as correct answer data.

13. The nuclear magnetic resonance imaging apparatus according to claim 12 , wherein the algorithm used for the machine learning includes a support vector machine or a decision tree.

14. The nuclear magnetic resonance imaging apparatus according to claim 12 or 13, wherein the supervised data includes extended supervised data generated by data extension.

15. The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor applies different calculation formulas depending on the movement of the subject when calculating the body movement information.

16. The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor outputs a classification result from the body motion information processing unit, and / or outputs an alert based on the classification result.

17. The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor determines whether or not to apply a body motion correction function based on a classification result from the body motion information processing unit.

18. The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor applies a body motion correction function based on the classification result from the body motion information processing unit, and outputs a body motion corrected image based on the body motion information.

19. The nuclear magnetic resonance imaging apparatus according to claim 1 , wherein the processor determines a plurality of regions for measuring motion information of the subject.

20. A body movement information processing method in which a body movement information processing device including a processor processes movement information of a subject placed in an imaging device, comprising: the processor: receiving a signal from a measurement device that measures motion information of the subject; Calculating body movement information from the signal; A body movement information processing method for classifying the body movement information.

Images