JP2012179252A - Medical image diagnostic device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a medical image diagnostic device to correctly calculate the utility of the dynamic heart image for one-cycle movements.SOLUTION: The medical image diagnostic device of an embodiment includes a structural information detector, a misalignment quantity calculator and a utility calculator. The structural information detector inputs the time-series dynamic image of a heart in cyclic movements and detects from this dynamic image, the structural information including at least one of the information such as the position, profile, long-axis angle or scale of the heart at a timing of specific time phase of the heart; the misalignment calculator calculates the misalignment quantity between the structural information at the specific time phase and the structural information in one cycle preceding or following the specific time phase; and the utility calculator calculates the degree of normalized utility of the misalignment quantity so that it can show if the dynamic heart image is useful in diagnosis.

Description

  Embodiments described herein relate generally to a medical image diagnostic apparatus.

  When diagnosing a periodically moving heart using an ultrasonic diagnostic apparatus, MRI, CT, or the like, a moving image consisting of a plurality of frames in one cycle is used.

  However, the image acquisition position between specific frames in the moving image for one period may be shifted due to the action of the subject during shooting. For example, the relative position of the ultrasonic probe of the ultrasonic diagnostic apparatus with respect to the body surface of the subject may be shifted during the acquisition of one cycle of moving images.

JP-A-11-128226 (refer to page 8, FIG. 9)

  However, there is a problem in that the usefulness of the obtained image cannot be correctly evaluated when the cross section of the heart is shifted during imaging.

  An object of the present invention is to provide a medical image diagnostic apparatus capable of correctly calculating usefulness in a moving image for one cycle of the heart.

  The medical image diagnostic apparatus according to the embodiment includes a structure information detection unit, a deviation amount calculation unit, and a usefulness calculation unit. The structure information detection unit inputs a time-series moving image obtained by imaging a heart that performs a periodic motion, and receives structure information including at least one of the position, contour, major axis angle, or scale of the heart from the moving image. , Detecting at the timing of the specific time phase of the heart. The deviation amount calculation unit calculates a deviation amount that is a difference between the structural information in the specific time phase and the structural information in the specific time phase one cycle before or after the specific time phase. The usefulness calculation unit calculates the usefulness obtained by normalizing the deviation amount so as to indicate whether or not the moving image is useful for diagnosis.

1 is a block diagram showing a configuration of a medical image diagnostic apparatus according to a first embodiment. The flowchart which shows operation | movement of 1st Embodiment. The block diagram which shows the structure of the medical image diagnostic apparatus concerning 2nd Embodiment. The flowchart which shows operation | movement of 2nd Embodiment. (A) is a diagram of an electrocardiogram waveform of the second embodiment, (b) is a diagram of an input moving image. The block diagram which shows the structure of the medical image diagnostic apparatus concerning 3rd Embodiment. The flowchart which shows operation | movement of 3rd Embodiment. The figure of the electrocardiogram waveform of 3rd Embodiment. The figure which shows the integration method of usefulness. (A) is a diagram of an electrocardiogram waveform of the third embodiment, (b) is a diagram of an input moving image.

  Hereinafter, a medical image diagnostic apparatus (hereinafter simply referred to as “diagnostic apparatus”) 10 according to an embodiment will be described with reference to the drawings.

  Hereinafter, the diagnostic apparatus 10 of the first embodiment will be described with reference to FIGS. 1 and 2.

  The configuration of the diagnostic apparatus 10 of this embodiment will be described with reference to FIG. FIG. 1 is a block diagram of the diagnostic apparatus 10.

  As illustrated in FIG. 1, the diagnostic device 10 includes a structure information detection unit 1, a deviation amount calculation unit 2, and a usefulness calculation unit 3.

  A time-series moving image of one cycle of the heart that performs periodic motion is input to the diagnostic apparatus 10. The moving image is input from a medical imaging apparatus such as an X-ray CT, MRI, or ultrasonic diagnostic apparatus, but may be input from an image server, a medium such as a CD or DVD, or a network storage.

  Since the inputted moving image is a moving image for one period, the structure information detection unit 1 sets the first and last images (frames) to a specific time phase, and sets the first frame (specific time before one period). Phase) and the structure information of the heart in the last frame (current specific time phase) is detected, and the structure information is output to the deviation amount calculation unit 2.

  The shift amount calculation unit 2 calculates the shift amount from the difference between the structure information of the input first frame (specific time phase one cycle before) and the structure information of the last frame (current specific time phase), The deviation amount is output to the usefulness calculation unit 3.

  The usefulness calculator 3 calculates the usefulness between the two specific time phases from the input deviation amount.

  The operation of the diagnostic apparatus 10 will be described with reference to FIGS. FIG. 2 is a flowchart showing the operation of the diagnostic apparatus 10.

  First, in step S1, the structure information detection unit 1 detects the structure information of the heart in the first frame and the last frame from the input moving image of one cycle of the heart that makes a periodic motion. The first frame and the last frame are images shifted by one cycle of the same specific time phase of the periodic motion of the heart.

  “Structural information” means the position, contour, posture, feature point, boundary, etc. of the heart. Specifically, the center position of the atrium, the center position of the ventricle, the center position of the aorta, the long axis angle of the heart, At least one of a scale, apex position, annulus position, or left ventricular boundary, left ventricular epicardial boundary, left atrial endocardial boundary, right ventricular boundary, right atrial endocardial boundary representing the outline of the heart Means. The “annular position” is the root of the mitral valve and the root of the aortic valve. In this embodiment, the case where the left ventricle center position, the major axis angle, and the scale are used as structure information, and the case where the left ventricular membrane boundary is used as structure information will be described. The scale indicates the length of an axis such as a major axis or a minor axis.

  When the structure information detection unit 1 detects structure information regarding the left ventricle center position, the major axis angle, and the scale, a feature amount cut out from an image around the left ventricle is used. As the feature amount, the luminance value or luminance gradient of the pixel is used. When detecting such structural information, the structural information detection unit 1 is described in Non-Patent Document 1 (Tatsuo Kozakaya, et al., “Fully Automatic Feature Localization for Medical Images using a Global Vector Concentration Approach” CVPR. 2008.). The discriminator is learned in advance, and the structural information is detected from the image using the discriminator.

  Also, structural information regarding the left ventricular membrane boundary is represented by a plurality of control points. When the structure information detection unit 1 detects this boundary, the active contour model described in Non-Patent Document 2 (TE Cootes, et al., “Active shape models, Their training and application” CVIU. 1995.). Is used. That is, the structure information detection unit 1 obtains the boundary by deforming the first shape using the contour shape model learned in advance and the luminance information around the boundary.

  Next, in step S <b> 2, the deviation amount calculation unit 2 determines the difference between the detected structural information of the first frame (specific time phase one cycle before) and the structural information of the last frame (current specific time phase). From this, the amount of deviation between the structural information detected in the two specific time phases is calculated.

For example, when the structural information is represented by the left ventricular center position C (x, y), the major axis angle θ, and the scale s, the deviation amount calculation unit 2 calculates the deviation amount D by Expression (1). Note that (x, y) is a coordinate value on a two-dimensional xy orthogonal coordinate axis set in the image.

When the structure information is represented by a point group related to the boundary control point X (x, y), the deviation amount calculation unit 2 calculates the deviation amount D by Expression (2).

  Note that the shift amount D is not limited to the sum of the square error of each point as shown in the equation (2), and for example, the difference between two pieces of structural information such as the sum of absolute value errors and the sum of distances between control points and boundaries. Any index may be used.

  Next, in step S3, the usefulness calculation unit 3 calculates the usefulness of the moving image between the two specific time phases based on the calculated deviation amount. Since the deviation amount is a physical amount and is difficult to understand intuitively, the deviation amount D is converted into a normalized value (usefulness) in order to notify the user (for example, a doctor).

  The usefulness calculation unit 3 gives a high usefulness if the shift amount D is small, and gives a low usefulness if it is large. The usefulness calculation unit 3 may use a conversion function prepared in advance for conversion from the deviation D to the usefulness, or use a table in which the value of the deviation D is associated with the usefulness. May be. The usefulness may be a continuous value, or may be a binary expression such as “white circle” or “useable” if a threshold value is set, and “x” or “not useable” if it is less.

For example, Expression (3) is used for the conversion function.

  However, E is a useful degree and D is a deviation | shift amount.

  In Formula (3), if the structure information matches, the deviation amount D is 0, and the maximum usefulness is E = 1. In addition, the usefulness decreases as the deviation D increases, and the usefulness E = 0 when not detected. In this way, the shift amount D can be normalized with 0 <= utility E <= 1.

  According to this embodiment, when the cross section of the heart being photographed is displaced due to an operator's operation mistake or the movement of the subject during photographing, the amount of displacement can be detected, and the input moving image is diagnosed. The user can be notified whether it is useful. That is, the diagnostic apparatus 10 can calculate the usefulness of diagnosis of a moving image for one period composed of a plurality of sheets.

  Hereinafter, the diagnostic apparatus 10 according to the second embodiment will be described with reference to FIGS.

  A configuration of the diagnostic apparatus 10 of the present embodiment will be described with reference to FIG. FIG. 3 is a block diagram showing the diagnostic apparatus 10 according to the present embodiment.

  The diagnostic device 10 includes a structure information detection unit 1, a deviation amount calculation unit 2, a usefulness calculation unit 3, a search range setting unit 5, a specific time phase detection unit 4, and an output unit 6.

  The specific time phase detection unit 4 detects the timing (time) of a specific time phase from image information or biological information (ECG, heart sound, pulse, etc.), and searches the structure information detection unit 1 for the timing of the specific time phase. Output to range setting unit 5.

  The structure information detection unit 1 is a search range setting unit that will later explain the structure information of the heart in the specific time phase detected by the specific time phase detection unit 4 from the input moving image as in the first embodiment. 5 is detected from the search range set in 5, and the structure information is output to the shift amount calculation unit 2 and the search range setting unit 5.

  The deviation amount calculation unit 2 calculates a deviation amount from the difference between the structure information of the current specific time phase and the structure information of the specific time phase one cycle before, and outputs the deviation amount to the usefulness calculation unit 3.

  The usefulness calculation unit 3 calculates the usefulness from the input deviation amount, and outputs the usefulness to the output unit 6.

  In the specific time phase detected by the specific time phase detection unit 4, the search range setting unit 5 uses the vicinity of the heart position detected by the structure information detection unit 1 as the search range of the specific time phase one cycle before. The search range is set and output to the structure information detection unit 1. Here, “the vicinity of the position” means a predetermined range with the position as the center or the reference point, and “the vicinity of the major axis angle” means the predetermined with the angle as the center. “A vicinity of the scale” means a predetermined range centered on the scale, and “a vicinity of the contour (eg, left ventricular membrane boundary)” means each of the boundaries. It means a range predetermined with the control point as a center or a reference point.

  The output unit 6 is a display such as a liquid crystal display device or a CRT, and displays usefulness, an electrocardiogram waveform, a moving image, and the like.

  Next, the operation of the diagnostic device 10 will be described with reference to FIGS. FIG. 4 is a flowchart showing the operation of the diagnostic apparatus 10. In the present embodiment, the case of the left ventricular membrane boundary will be described as “structure information”. The left ventricular membrane boundary is represented by a plurality of control points.

  First, in step S <b> 11, the specific time phase detection unit 4 detects the timing of the specific time phase of the heart that repeats the input periodic motion, and outputs it to the structure information detection unit 1.

  “Specific time phase” means the timing of the left ventricular end diastole or the left ventricular end systole, which is a time phase that can be estimated from the electrocardiograph, and from the time phase that can be estimated from the electrocardiograph It is a time phase advanced for a fixed time. In the present embodiment, a case will be described in which the specific time phase is the time phase at the end diastole of the left ventricle. The specific time phase may be estimated from a moving image of the heart instead of the electrocardiograph.

  Next, in step S <b> 12, the search range setting unit 5 sets a search range for detecting heart structure information and outputs the search range to the structure information detection unit 1. That is, the search range setting unit 5 determines the first search range related to the structure information of the current specific time phase (end diastole), and searches for the structure information of the specific time phase one cycle before (end diastole one cycle ago). Determine the range.

  In addition, the search range setting part 5 sets the area | region preset as a 1st search range, when searching for the first time.

In addition, when the left ventricular boundary is detected in the specific time phase at the end diastole one cycle before, the search range setting unit 5 is in the vicinity of each control point of the boundary and from the first search range. A narrow area is set as the second search range. For example, the second search range is set as shown in Equation (4).

  Further, when the left ventricular membrane boundary is not detected in the specific time phase at the end diastole one cycle before, the search range setting unit 5 sets a preset region as the second search range.

  As described above, by narrowing the second search range one cycle before the current specific time phase to be smaller than the first search range, it is possible to contribute to reduction in search processing amount and detection error.

  Next, in step S <b> 13, as described in the first embodiment, the structure information detection unit 1 detects the heart structure information from the set search range and outputs it to the deviation amount calculation unit 2.

  Next, in step S14, the deviation amount calculation unit 2 calculates the difference between the detected left ventricular boundary at the specific time phase at the end diastole and the left ventricular boundary at the specific time phase at the end diastole one cycle before. The shift amount D of the structural information between the two specific time phases is calculated and output to the usefulness calculation unit 3.

  The calculation method of the shift amount D is the same as that in step S2 of the first embodiment. However, if the left ventricular membrane boundary is not detected from the set search range, the shift amount calculation unit 2 determines that the shift amount calculation unit 2 is not detected. As a result, the shift amount D is set to infinity.

  Next, in step S <b> 15, the usefulness calculation unit 3 calculates the usefulness between the two specific time phases based on the deviation amount D, and outputs the usefulness to the output unit 6. The method for calculating the usefulness is the same as that in step S3 of the first embodiment, and thus description thereof is omitted.

  Finally, in step S16, the output unit 6 notifies the user of the calculated usefulness.

  The output unit 6 displays the image shown in FIG. The electrocardiogram waveform representing the heart motion cycle shown in FIG. 5A and the usefulness calculated thereon are displayed in a superimposed manner. When the usefulness is high, it is represented by a white circle, and when it is low, it is indicated by x. Yes. FIG. 5B shows the input moving image. Further, the output unit 6 may display the electrocardiographic waveform of each period in different colors according to the usefulness.

  According to this embodiment, when the cross section of the heart being photographed is displaced due to an operator's operation mistake or the movement of the subject during photographing, the amount of deviation can be detected, and the input moving image is used for diagnosis. Users can be notified if it is useful. Further, by narrowing the search range in the specific time phase one cycle before using the current structure information of the specific time phase, it is possible to reduce search processing time and detection errors.

  Hereinafter, a diagnostic apparatus 10 according to a third embodiment will be described with reference to FIGS.

  The configuration of the diagnostic device 10 of this embodiment will be described with reference to FIG. FIG. 6 is a block diagram showing the diagnostic device 10.

  As shown in FIG. 6, the present embodiment has a configuration in which a connection between the specific time phase detection unit 4 and the deviation amount calculation unit 2 of the diagnostic device 10 of the second embodiment is added. With this configuration, the shift amount calculation unit 2 can determine the specific time phase of the currently focused frame. Therefore, even when a plurality of sets of specific time phases are set in one cycle, it is possible to calculate the shift amount of the structure information in each previous cycle. Since other configurations are the same as those of the second embodiment, description thereof is omitted.

  Next, the operation of the diagnostic device 10 will be described with reference to FIGS. FIG. 7 is a flowchart showing the operation of the diagnostic apparatus 10.

  First, in step S21, the specific time phase detection unit 4 detects two or more sets of specific time phase timings of the heart that repeats the input periodic motion from one cycle, and the structure information detection unit 1, the deviation amount calculation unit Output to 2.

  In the present embodiment, a case will be described in which two specific time phases A and B are selected from one period as shown in FIG.

  Next, in step S <b> 22, the search range setting unit 5 sets a search range for detecting heart structure information and outputs the search range to the structure information detection unit 1.

  Next, in step S <b> 23, the structure information detection unit 1 detects the heart structure information from the set search range and outputs it to the deviation amount calculation unit 2.

  Steps S22 and S23 are the same as the operations in steps S12 and S13 in the second embodiment, and thus description thereof is omitted.

  Next, in step 24, the deviation amount calculation unit 2 determines the left ventricular membrane boundary in the current specific time phase and the specific time phase one cycle before (if the currently focused frame is the specific time phase A2 in FIG. For example, from the left ventricular membrane boundary in the specific time phase A1 and the specific time phase B2 in the case of the specific time phase B2, the deviation amounts of the structural information between the two sets of specific time phases are calculated, and the usefulness calculating unit 3 Output. Since the calculation method of each deviation amount is the same as that in step S2 of the first embodiment, description thereof is omitted.

  Next, in step S25, the usefulness calculation unit 3 calculates the usefulness between the two sets of specific time phases based on the respective deviation amounts. Step S25 is the same as the operation of step S15 in the second embodiment, and a description thereof will be omitted.

  Next, in step S ′ 25, the usefulness degree calculation unit 3 integrates the usefulness levels between the two sets of specific time phases and outputs them to the output unit 6.

  For example, the usefulness is integrated as shown in FIG. Specifically, when the usefulness calculation unit 3 has a white circle between A2 and A3, if either one between B2 and B2 partially overlapping with A2 or A3 or between B2 and B3 is X, The integrated usefulness is calculated as x (see FIG. 9A), and if both are white circles, the integrated usefulness is calculated as a white circle (see FIG. 9B). When the usefulness is a continuous value, the usefulness calculating unit 3 adopts a value having a low usefulness. That is, the usefulness calculation unit 3 calculates the usefulness between the specific time phases of each group, and selects the lowest usefulness among the usefulness of each group.

  In this way, by setting two sets of specific time phases in one cycle and integrating their usefulness, even if the cross section is once shifted in one cycle and then returned to the same position, Deviation can be detected.

  Finally, in step S26, the output unit 6 notifies the user of the calculated usefulness. The output unit 6 displays an image as shown in FIG. FIG. 10 (a) displays an electrocardiogram waveform representing a cardiac motion cycle, usefulness between specific time phases A and B calculated in the same, and integrated usefulness in an overlapping manner. Is. FIG. 10B shows the input moving image. Further, the output unit 6 may display the electrocardiographic waveform of each period in different colors according to the usefulness.

  According to the present embodiment, it is possible to detect the displacement of the cross section of the heart during imaging more accurately, and to realize more robust evaluation of moving images.

  In addition, in the said embodiment, although demonstrated between two sets of specific time phases, it is not restricted to this, You may obtain | require a usefulness between three or more sets of specific time phases.

Example of change

  In each of the embodiments described above, the shift amount D is calculated by the difference between the structure information of an arbitrary specific time phase in the moving image and the structure information of the specific time phase one cycle before the arbitrary specific time phase. However, when the moving image is recorded, the shift amount D includes the structure information of an arbitrary specific time phase in the moving image and the structure information of a specific time phase after one cycle of the arbitrary specific time phase. The difference may be calculated.

  The diagnostic device 10 of each of the above embodiments can be realized by using, for example, a general-purpose computer as basic hardware. That is, the structure information detection unit 1, the deviation amount calculation unit 2, the usefulness calculation unit 3, the specific time phase detection unit 4, and the search range setting unit 5 cause the processor mounted on the computer to execute the program. Can be realized. At this time, the diagnostic device 10 may be realized by installing the above program in a computer in advance, or may be stored in a storage medium such as a CD-ROM or distributed through the network. This program may be realized by appropriately installing it on a computer. B and C can be realized by appropriately using a memory, a hard disk, or a storage medium such as a CD-R, a CD-RW, a DVD-RAM, a DVD-R, or the like that is built in or externally attached to the computer. it can.

  Although one embodiment of the present invention has been described above, this embodiment is presented as an example and is not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 1 ... Structural information detection part, 2 ... Deviation amount calculation part, 3 ... Usefulness calculation part

Claims (9)

A time-series moving image obtained by imaging a periodically moving heart is input, and structural information including at least one of the position, contour, major axis angle, or scale of the heart is specified from the moving image. A structure information detector that detects at the timing of the time phase;
A deviation amount calculation unit that calculates a deviation amount that is a difference between the structure information in the specific time phase and the structure information in the specific time phase one cycle before or after the specific time phase;
A usefulness calculating unit that calculates a usefulness obtained by normalizing the deviation amount, so as to indicate whether the moving image is useful for diagnosis;
A medical image diagnostic apparatus comprising: A first search range for detecting the structure information is set in the vicinity of the position, contour, major axis angle, or scale of the structure information in the specific time phase, and one cycle before the specific time phase, Or a search range setting unit that sets the second search range in the specific time phase after the one cycle to a region narrower than the first search range,
The medical image diagnostic apparatus according to claim 1. An electrocardiograph that measures the electrocardiogram of the heart, or a time phase detection unit that detects the specific time phase from the moving image;
The medical image diagnostic apparatus according to claim 1. When the structure information detection unit cannot detect the structure information in the specific time phase before the one cycle or after the one cycle, the shift amount calculation unit sets the shift amount to infinity. ,
The medical image diagnostic apparatus according to claim 1. When the structure information detection unit detects the structure information for the first time in the specific time phase, or when it has not been detected, the search range setting unit, the preset region as the search range,
The medical image diagnostic apparatus according to claim 2. The medical image diagnostic apparatus is an ultrasonic diagnostic apparatus having an image acquisition unit that acquires the moving image using an ultrasonic probe.
The medical image diagnostic apparatus according to claim 1. The specific time phase detection unit detects the specific time phase of a plurality of sets,
The structure information detection unit calculates the structure information between the specific time phases of each set,
The deviation amount calculation unit calculates the deviation amount of the structure information detected between the specific time phases of the sets,
The usefulness calculation unit calculates the usefulness between the specific time phases of each set, and selects the lowest usefulness among the usefulness of each set.
The medical image diagnostic apparatus according to claim 1. The position of the structural information is the apex position of the heart, the center position of the atrium, the center position of the ventricle, the center position of the aorta, or the annulus position, and the contour is the left ventricular membrane of the heart Border, left ventricular epicardial border, left atrial endocardial border, right ventricular border, or right atrial endocardial border,
The medical image diagnostic apparatus according to claim 1. An output unit for notifying the user of the usefulness;
The medical image diagnostic apparatus according to claim 1.

Images