Diagnostic support program

Abstract

To provide a diagnostic support program that is possible to display a movement of an area whose shape changes for each respiratory element including all or a part of expired air or inspired air.SOLUTION: The diagnostic support program includes processing of acquiring a plurality of frame images from a database, processing of specifying a cycle of a respiratory element including all or a part of expired air or inspired air based on pixels in a specific area in each of the frame images, processing of detecting a lung field based on the cycle of the specified respiratory element, processing of dividing the detected lung field into a plurality of block areas and calculating a change in an image in a block area in each of the frame images, processing of Fourier-transforming a change in the image in each block area in each of the frame images, processing of extracting a spectrum in a fixed band including a spectrum corresponding to the cycle of the respiratory element, out of spectra obtained after the Fourier-transforming, processing of performing inverse Fourier transform on the spectrum extracted from the fixed band, and processing of displaying each of the images after performing the inverse Fourier transform, on a display.SELECTED DRAWING: Figure 6A

Description

Technical Field

[0001] The present invention relates to a technique for analyzing a human body image and displaying an analysis result.

Background Art

[0002] When a doctor diagnoses the lungs using dynamic images of the chest, it is important to observe a time series of dynamic chest images in which the subject is photographed in a natural breathing state. A spirometer that can easily acquire physiological data, a RI (Radio Isotope) test, a simple X-ray photograph that can obtain morphological data, CT (Computed Tomography), etc. are known as methods for evaluating lung function. However, it is not easy to efficiently acquire both physiological data and morphological data.

[0003] In recent years, a method has been attempted in which a dynamic image of the human chest is photographed using a semiconductor image sensor such as an FPD (Flat panel detector) and used for diagnosis. For example, Non-Patent Document 1 discloses a technique in which a difference image indicating a difference in signal values is generated between a plurality of frame images constituting a dynamic image, and the maximum value of each signal value is obtained from the difference image and displayed.

[0004] Also, Patent Document 1 discloses a technique in which a lung field region is extracted from each of a plurality of frame images showing the dynamics of the human chest, the lung field region is divided into a plurality of small regions, and the divided small regions are associated with each other and analyzed between the plurality of frame images. According to this technique, a feature amount indicating the movement of the divided small region is displayed.

Prior Art Documents

Patent Documents

[0005] [[ID=3(continue)]]

Patent Document 1

Non-Patent Documents

[0006] [Non-Patent Document 1] “Basic Imaging Properties of a Large Image Intensifier-TV Digital Chest Radiographic System” Investigative Radiology:April 1987; 22: 328-335. [Overview of the Initiative] [Problems that the invention aims to solve]

[0007] However, simply displaying the maximum value of the inter-frame difference for each pixel of a dynamic image, as in the technology described in Non-Patent Document 1, does not make it easy for a physician to understand the pathological condition. Similarly, simply displaying feature quantities, as in the technology described in Patent Document 1, is also insufficient for understanding the pathological condition. Therefore, it is desirable to display images that correspond to the state of respiration and pulmonary blood vessels. In other words, it is desirable to understand the respiratory state and overall vascular dynamics of the human body being studied, and to display images that show the actual movement based on the waveform or frequency of respiration, heart, blood vessels in the pulmonary hilum, or blood flow, or the trend of changes in the image.

[0008] This invention has been made in view of these circumstances and aims to provide a diagnostic support program capable of displaying the movement of regions whose shape changes for each respiratory element, including all or part of exhalation or inhalation. More specifically, it aims to calculate numerical values ​​that assist in diagnosis by quantifying the agreement rate and other discrepancies with the waveform shape and Hz already acquired for data of a new target to be measured, and further generate images that assist in diagnosis by imaging these numerical values. [Means for solving the problem]

[0009] (1) In order to achieve the above objective, the present invention employs the following means. Specifically, a diagnostic support program according to one aspect of the present invention is a diagnostic support program that analyzes images of the human body and displays the analysis results, characterized in that it causes a computer to perform the following: a process of acquiring a plurality of frame images from a database storing the images; a process of identifying at least one frequency of a respiratory element including all or part of exhalation or inhalation based on pixels of a specific region of each frame image; a process of detecting a lung field based on at least one frequency of the identified respiratory element; a process of dividing the detected lung field into a plurality of block regions and calculating the image changes of the block regions in each frame image; a process of performing a Fourier transform on the image changes of each block region in each frame image; a process of extracting a spectrum within a certain band from the spectrum obtained after the Fourier transform that includes a spectrum corresponding to at least one frequency of the respiratory element; a process of performing an inverse Fourier transform on the spectrum extracted from the certain band; and a process of displaying each image after the inverse Fourier transform on a display.

[0010] (2) A diagnostic support program according to one aspect of the present invention further includes a process of using a filter to extract, from the spectrum obtained after the Fourier transform, a spectrum within a certain band that includes the frequency of noise and frequencies other than the frequency of the respiratory element obtained from the frame image, or a spectrum corresponding to the input frequency or frequency band.

[0011] (3) A diagnostic support program according to one aspect of the present invention further includes a process of generating an image between frames based on the frequency of the respiratory element and each of the frame images.

[0012] (4) A diagnostic support program according to one aspect of the present invention is a diagnostic support program that analyzes images of a human body and displays the analysis results, characterized in that it causes a computer to perform the following processes: acquiring a plurality of frame images from a database storing the images; identifying at least one frequency of a cardiovascular beat element extracted from the heartbeat or vascular beat of a subject; identifying at least one frequency of a respiratory element including all or part of exhalation or inhalation based on pixels in a specific region of each frame image; detecting a lung field based on the at least one frequency of the identified respiratory element; dividing the detected lung field into a plurality of block regions and calculating the image changes of the block regions in each frame image; performing a Fourier transform on the image changes of each block region in each frame image; extracting a spectrum within a certain band from the spectrum obtained after the Fourier transform that includes a spectrum corresponding to at least one frequency of the cardiovascular beat element; performing an inverse Fourier transform on the spectrum extracted from the certain band; and displaying each image after the inverse Fourier transform on a display.

[0013] (5) Furthermore, a diagnostic support program according to one aspect of the present invention is a diagnostic support program that analyzes images of a human body and displays the analysis results, characterized in that it causes a computer to perform the following processes: acquiring a plurality of frame images from a database storing the images; identifying at least one frequency of a cardiovascular beat element extracted from the heartbeat or vascular beat of a subject; detecting a lung field; dividing the detected lung field into a plurality of block regions and calculating the image changes of the block regions in each frame image; performing a Fourier transform on the image changes of each block region in each frame image; extracting a spectrum within a certain band from the spectrum obtained after the Fourier transform that includes a spectrum corresponding to at least one frequency of the cardiovascular beat element; performing an inverse Fourier transform on the spectrum extracted from the certain band; and displaying each image after the inverse Fourier transform on a display.

[0014] (6) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that it further includes a process of using a filter to extract from the spectrum obtained after the Fourier transform a spectrum that includes the frequency of noise and frequencies other than the frequencies of cardiovascular beat elements obtained from the frame image, or a spectrum within a certain band that includes the input frequency or frequency band.

[0015] (7) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that it further includes a process for generating an image between frames based on the frequencies of the identified cardiovascular rhythm elements and the respective frame images.

[0016] (8) Furthermore, a diagnostic support program according to one aspect of the present invention is a diagnostic support program that analyzes images of a human body and displays the analysis results, characterized in that it causes a computer to perform the following processes: acquiring a plurality of frame images from a database storing the images; identifying at least one frequency of a vascular pulse element extracted from the vascular pulse of a subject; dividing the analysis range set for each frame image into a plurality of block regions and calculating the image changes of the block regions in each frame image; performing a Fourier transform on the image changes of each block region in each frame image; extracting a spectrum within a certain band from the spectrum obtained after the Fourier transform that includes a spectrum corresponding to at least one frequency of the vascular pulse element; performing an inverse Fourier transform on the spectrum extracted from the certain band; and displaying each image after the inverse Fourier transform on a display.

[0017] (9) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that it further includes a process of using a filter to extract from the spectrum obtained after the Fourier transform a spectrum that includes the frequency of noise and frequencies other than the frequency of the vascular pulse element obtained from the frame image, or a spectrum within a certain band that includes the input frequency or frequency band.

[0018] (10) A diagnostic support program according to one aspect of the present invention further includes a process for generating an image between frames based on the frequencies of the identified cardiovascular beat elements and the respective frame images.

[0019] (11) A diagnostic support program according to one aspect of the present invention is a diagnostic support program that analyzes images of the human body and displays the analysis results, characterized in that it causes a computer to perform the following: a process of acquiring a plurality of frame images from a database storing the images; a process of identifying at least one frequency of a respiratory element including all or part of exhalation or inhalation based on the pixels of a specific region of each frame image; a process of detecting the lung field and diaphragm based on the frequency of at least one of the identified respiratory elements; a process of dividing the detected lung field into a plurality of block regions and calculating the rate of change of pixels in each of the frame images of the block region; a process of extracting only block regions in which the synchronization rate is within a predetermined range using a synchronization rate which is the ratio of the rate of change of pixels in the block region to the rate of change of a dynamic region linked to respiration; and a process of displaying each image containing only the extracted block regions on a display.

[0020] (12) A diagnostic support program according to one aspect of the present invention further includes a process to identify at least one frequency of a cardiovascular beat element extracted from the heartbeat or vascular beat of a subject, or at least one frequency of a vascular beat element extracted from a vascular beat.

[0021] (13) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that the logarithmic value of the synchronization rate is set to a certain range including 0.

[0022] (14) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by further including a process to detect lung fields in other frames using at least one Bezier curve on a lung field detected in a specific frame.

[0023] (15) Further, the diagnostic support program according to one aspect of the present invention is characterized in that it selects internal control points within the detected lung field and divides the lung field by a curve or a straight line passing through the internal control points within the lung field.

[0024] (16) Further, the diagnostic support program according to one aspect of the present invention is characterized in that it relatively increases the interval between control points in the outer extension of the detected lung field and its vicinity, and relatively decreases the interval between the internal control points according to the inflation ratio for each part within the detected lung field.

[0025] (17) Further, the diagnostic support program according to one aspect of the present invention is characterized in that, within the detected lung field, it relatively increases the interval between control points as it progresses in the cranio-caudal direction with respect to the human body, or relatively increases the interval according to a specific vector direction.

[0026] (18) Further, the diagnostic support program according to one aspect of the present invention further includes a process of detecting the lung field in other frames using at least one or more Bezier surfaces on the lung field detected in a specific frame.

[0027] (19) Further, the diagnostic support program according to one aspect of the present invention further includes a process of detecting a range corresponding to the analysis range in other frames using at least one or more Bezier curves on a predetermined analysis range in a specific frame.

[0028] (20) Further, the diagnostic support program according to one aspect of the present invention further includes a process of drawing at least the lung field, blood vessels, or the heart using at least one or more Bezier curves.

[0029] (21) Furthermore, a diagnostic support program according to one aspect of the present invention is a diagnostic support program that analyzes images of the human body and displays the analysis results, characterized in that it causes a computer to perform the following processes: acquiring a plurality of frame images from a database storing the images; identifying an analysis range using a Bézier curve for all of the acquired frame images; and detecting an analysis target based on the change in intensity within the analysis range.

[0030] (22) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by further including a process for calculating the characteristics of the edges of the detected object to be analyzed.

[0031] (23) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that it detects the diaphragm by calculating the difference in intensity for each consecutive image, and displays an index indicating the position or shape of the detected diaphragm or a dynamic part linked to respiration.

[0032] (24) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by displaying the diaphragm that is not obstructed by parts other than the diaphragm by changing the intensity threshold, and interpolating the overall shape of the diaphragm.

[0033] (25) A diagnostic support program according to one aspect of the present invention further includes a process of calculating the frequency of at least one respiratory element from the detected position or shape of the diaphragm or the position or shape of a dynamic part linked to respiration.

[0034] (26) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by further including a process to spatially normalize the detected lung field or to temporally normalize it using reconstruction.

[0035] (27) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by correcting the respiratory element by changing the phase of at least one frequency of the respiratory element or by smoothing the waveform of the respiratory element.

[0036] (28) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by identifying a waveform of any part within the analysis range, extracting the frequency components of the identified waveform, and outputting an image corresponding to the frequency components of the waveform.

[0037] (29) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by detecting the density of the analysis range and removing areas where the density changes relatively significantly.

[0038] (30) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by further including a process of selecting at least one frequency to perform an inverse Fourier transform from the spectrum obtained after the Fourier transform, based on the spectral composition ratio of organ-specific periodic changes.

[0039] (31) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that it controls an X-ray imaging device to adjust the X-ray irradiation interval according to the frequency of at least one of the respiratory elements.

[0040] (32) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that, after the inverse Fourier transform, it extracts and displays only the blocks with relatively large amplitude values.

[0041] (33) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that, after identifying the lung field, it further includes a process of identifying the diaphragm or thoracic cage, calculating the amount of change in the diaphragm or thoracic cage, and calculating the rate of change from the amount of change.

[0042] (34) A diagnostic support program according to one aspect of the present invention further includes a process of multiplying a specific spectrum by a coefficient, and is characterized by performing highlighting based on the specific spectrum multiplied by the coefficient.

[0043] (35) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that, after acquiring multiple frame images from a database storing images, it applies a digital filter to the area to be analyzed in order to identify the frequency or waveform of respiratory elements.

[0044] (36) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized in that it identifies a plurality of frequencies of respiratory elements, including all or part of exhalation or inhalation, based on the pixels of a specific region of each frame image, and displays each image corresponding to each of the plurality of frequencies of the respiratory elements on a display.

[0045] (37) Furthermore, a diagnostic support program according to one aspect of the present invention is characterized by selecting images that cluster together at a certain value within a specific range of one or more frame images and displaying them on a display. [Effects of the Invention]

[0046] According to one aspect of the present invention, it is possible to display the movement of regions whose shape changes for each respiratory element, including all or part of exhalation or inhalation. [Brief explanation of the drawing]

[0047] [Figure 1A] This figure shows the schematic configuration of the diagnostic support system according to this embodiment. [Figure 1B] This figure shows an example of a method for segmenting the lung region. [Figure 1C] This diagram shows how the morphology of the lungs changes over time. [Figure 1D] This diagram shows how the morphology of the lungs changes over time. [Figure 2A]This figure shows the change in "intensity" of a specific block and the results of Fourier analysis. [Figure 2B] This figure shows the Fourier transform result of extracting frequency components close to the heartbeat, and the inverse Fourier transform of this result, which shows the change in "intensity" of the frequency components close to the heartbeat. [Figure 2C] This figure shows an example of extracting a specific frequency band from the spectrum obtained after the Fourier transform. [Figure 2D] This diagram schematically represents the rate of change in the lungs. [Figure 2E] This figure shows an example of a pattern image of the lung field region. [Figure 2F] This figure shows an example of a pattern image of the lung field region. [Figure 2G] This figure shows an example of a pattern image of the lung field region. [Figure 2H] This figure shows an example of a pattern image of the lung field region. [Figure 3A] This figure shows an example of plotting the contours of the lung fields using both Bézier curves and straight lines, representing the state where the lung fields are at their maximum size. [Figure 3B] This figure shows an example of plotting the contours of the lung fields using both Bézier curves and straight lines, representing the state where the lung fields are at their smallest. [Figure 4A] This image shows the front and back of the lung field images superimposed between the previous and next frames. [Figure 4B] Figure 4A shows the result of taking the difference between the two original images, resulting in a state where a "line with a strong gap" is created. [Figure 4C] Figure 4B shows the difference in the sum of the "intensity" values ​​and "density" values ​​at each position in the vertical direction of the image. [Figure 5] This figure shows the result of approximating the relative position of the diaphragm using curve regression. [Figure 6A] This flowchart shows an overview of the respiratory function analysis according to this embodiment. [Figure 6B] This figure shows an example of an image displayed on a screen. [Figure 6C]This figure shows an example of an image displayed on a screen. [Figure 7] This flowchart shows an overview of the pulmonary blood flow analysis according to this embodiment. [Figure 8] This flowchart outlines other blood flow analyses according to this embodiment. [Figure 9] This figure shows an example of multiplying a certain spectrum obtained after a Fourier transform by a coefficient. [Figure 10] This figure shows an example of a lung field plotted using Bézier curves. [Figure 11] This figure shows an example of dividing the lung field using Bézier curves. [Figure 12] This figure shows an example of dividing the lung field using Bézier curves. [Figure 13] This figure shows an example of comparing the waveforms of aortic blood flow and ventricular volume. [Figure 14] This figure shows an example of pixel values ​​for the lungs and the area surrounding them. [Figure 15] This is a schematic diagram illustrating the general structure of blood vessels in the human body. [Modes for carrying out the invention]

[0048] First, let me explain the basic concept of this invention. In this invention, movements in the human body such as respiration, blood vessels, lung area and volume, and other biological movements that are perceived as repeating at a certain period, are captured as waves, either as a whole or as a partial range, representing a certain repetition or routine on the time axis, and measured. The measurement results of the waves are either (a) the wave shape itself, or (b) the interval between waves (frequency: Hz). These two concepts together are called "base data".

[0049] It is possible for waves to be linked in the same way at the same time. For example, in the case of breathing, the following approximation can be conceived. (Average change in "density" over a rough range) ≈ (Changes in the rib cage) ≈ (Movement of the diaphragm) ≈ (Lung function test) ≈ (Chora-abdominal respiratory sensor)

[0050] Regarding "(a) the wave form itself" above, the concept of "waveform tunability" is used, and the image is displayed based on this (Wave form tunable imaging). Also, regarding "(b) the wave interval (frequency: Hz)" above, the concept of "frequency tunability" is used, and the image is displayed based on this (Frequency tunable imaging).

[0051] For example, in the case of the heart, as shown in Figure 13, "An example comparing the waveforms of aortic blood flow and ventricular volume," the peak of aortic blood flow and the peak or waveform of ventricular volume do not coincide. However, if we define one cycle as a time interval of equal intervals, such as from time t1 to t2, from time t2 to t3, from time t3 to t4, etc., then one cycle of aortic blood flow and one cycle of ventricular volume will be repeated many times, and it can be said that the frequencies of each waveform are synchronized. Focusing on these waveforms, by identifying one cycle from measured values ​​as shown in Figure 13 and using a model waveform, it is possible to predict the shape of the wave (wave form). In other words, as a way to create the "waveform as base data," it can be done by measurement, generated from frequency (cycle), using a model waveform, or by averaging waveforms between individuals. If the cycle (period) of organs with frequencies, such as the heart, is known, the wave shape can be predicted. This allows for the understanding of waveforms such as aortic blood flow and ventricular volume, and based on these waveforms, it becomes possible to display dynamic images of the organs.

[0052] Furthermore, when acquiring changes in "density" such as respiration, heart rate, and lung hilum, it is advisable to apply a digital filter beforehand to prevent other elements from being mixed in.

[0053] Furthermore, this invention uses the concept of a "respiratory element." A "respiratory element" includes all or part of exhalation or inhalation. For example, "one breath" can be considered as being divided into "one exhalation" and "one inhalation," or it can be limited to one of "0%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100%" of "one exhalation or one inhalation." It is also possible to extract and evaluate only a certain percentage of each exhalation, for example, only 10% of the exhalation. Using any of these data, or a combination thereof, it is possible to extract images with higher accuracy. In this process, calculations may be performed repeatedly and mutually.

[0054] This way of thinking can be applied not only to the "respiratory element" but also to the "cardiovascular element."

[0055] Here, when creating the base data, features obtained from one or more modalities (for example, a change amount composed of "density" and "volumetry" within a certain range, thoracic movement, diaphragmatic movement, "spirometry," and two or more thoracoabdominal respiratory sensors), or multiple waveform measurements such as the same respiratory cycle, complement each other's component extraction to improve accuracy. This makes it possible to reduce artifacts and improve accuracy based on certain predictions such as lines. Here, "density" is translated as "density," but in imaging, it means the "absorption value" of pixels in a specific region. For example, in CT, air is used as "-1000," bone as "1000," and water as "0."

[0056] Furthermore, the waveform's axis, width, range, and Hz fluctuations and amplitude are estimated by extracting components from each other. In other words, through multiple superpositions, the Hz axis setting is averaged, and the optimal range of the axis, width, range, and Hz is calculated through variance. In this process, Hz (noise) from other actions may be extracted, and the relative extent to which that wave is excluded may also be measured. That is, in some cases, only a portion of the waveform elements may be extracted from the entire waveform.

[0057] In this specification, "density" and "intensity" are used distinctly. "Density," as described above, refers to the absorption value. In the source image of XP or XP video, air has high permeability, and areas with high permeability are represented as white, with air being "-1000," water "0," and bone "1000." On the other hand, "intensity" is a relative change from "density," for example, a normalized value that "converts" it into a range of density or signal intensity. In other words, "intensity" is a relative value in an image, such as brightness or emphasis. While directly dealing with the absorption value of an XP image, it is expressed as "density" or "change in density (Δdensity)." Then, for the sake of image representation, this is converted as described above and expressed as "intensity." For example, when displaying color in 256 gradations from 0 to 255, it becomes "intensity." This distinction in terminology applies to XP and CT.

[0058] On the other hand, in the case of MRI, even if one tries to define air as "-1000," water as "0," and bone as "1000," the values ​​change significantly depending on the MRI pixel values, the type of measuring machine, the person's physical condition and build at the time of measurement, and the measurement time. Furthermore, the way MRI signals are acquired, such as T1-weighted images, varies depending on the facility and the type of measuring machine, and is not consistent. For this reason, in the case of MRI, a "density" definition like that used for XP or CT is not possible. Therefore, in MRI, relative values ​​are handled from the initial imaging stage, and are expressed as "intensity" from the beginning. The signal processed is also "intensity."

[0059] Based on the above, it becomes possible to obtain base data. For the new target to be measured, the waveform and Hz of the base data are extracted within a certain width and range. For example, extraction can be limited to respiratory data only, or to a width, range, and waveform element similar to that of vascular data. The width of this waveform and Hz is determined relatively and statistically based on factors such as waveform elements from other functions, "artifacts" like noise, waveforms of other "modalities" that appear to have synchronicity, and reproducibility obtained through multiple trials. Adjustment and experience are necessary here (machine learning can also be applied). This is because widening the width and range introduces elements from other functions, while narrowing it too much removes elements from the function itself; therefore, adjustment of the range is necessary. For example, having data from multiple trials makes it easier to define the range, Hz, and measurement agreement width.

[0060] [Regarding the rate of agreement / consensus] In this specification, the trend of image changes is described as the synchronization rate. For example, the lung field is detected and divided into multiple block regions, and the "average density (pixel value x)" of the block region in each frame image is calculated. Then, the ratio (x') of the average pixel value of the block region in each frame image to the range of change from the minimum to the maximum value of the "average density (pixel value x)" (0% to 100%) is calculated. On the other hand, using the ratio (x' / y') of the ratio (y') of the change in the diaphragm (y) in each frame image to the range of change from the minimum to the maximum position of the diaphragm (0% to 100%), only block regions in which the ratio (x' / y') falls within a predetermined range are extracted.

[0061] Here, if y'=x' or y=ax (where a is the numerical value of the diaphragm amplitude or the coefficient of the "density" value), it is a perfect match. However, not only perfect matches are meaningful values; values ​​within a certain range should be extracted. Therefore, in one aspect of the present invention, a certain range is defined using logarithms (log) as follows. That is, when calculated as the percentage of y=x, a perfect match in tuning is "log y' / x'=0". Furthermore, when extracting a range of tuning agreement rates that is narrow (mathematically narrow), for example, a range close to 0 is defined as "log y' / x'=-0.05 to +0.05", and when extracting a range of tuning agreement rates that is wide (mathematically wide), for example, a range close to 0 is defined as "log y' / x'=-0.5 to +0.5". In other words, the logarithmic value of the tuning agreement rate is defined as a certain range including 0. The narrower this range, and the higher the number of matching values ​​within that range, the higher the matching rate can be said to be.

[0062] When this ratio is calculated for each pixel and the number of occurrences is counted, a normal distribution is obtained for healthy individuals, with a peak at the case of perfect match. In contrast, for individuals with disease, the distribution of this ratio becomes distorted. It should be noted that the method of determining the range using logarithms, as described above, is merely one example, and the present invention is not limited to this. That is, the present invention performs "image extraction" by assuming that (change in "density" within a certain rough range) ≈ (change in the rib cage) ≈ (movement of the diaphragm) ≈ (pulmonary function test) ≈ (movement of the thoracoabdominal respiratory sensor) ≈ (area and volume of the lung field), and methods other than those using logarithms can also be applied. By such methods, it becomes possible to display synchronized images.

[0063] In the case of blood vessels, a series of changes in "density" (x (a single waveform at the pulmonary hilum)) that occur in response to a series of cardiac contractions (y) have a slight time delay (phase change), so they can be expressed as y=a'(xt) (i.e., Y≒X). In the case of perfect match, t=0, so y=x or y=a'x. Similar to the case of the diaphragm, when extracting a range of synchronization rate that is narrow (mathematically narrow), for example, a range close to 0 is defined as "log y' / x'=-0.05 to +0.05", and if the range of synchronization rate is wide (mathematically wide), for example, a range close to 0 is defined as "log y' / x'=-0.5 to +0.5". The narrower this range is, and the higher the numerical values ​​that match within that range, the higher the synchronization rate can be said to be.

[0064] For other blood vessels, the "portion corresponding to the heart" mentioned above is excluded, and the "density" on the central side plotted from the pulmonary hilum is used. Peripheral blood vessels can be treated similarly.

[0065] Furthermore, the present invention can also be applied to the circulatory system. For example, a change in cardiac "density" is directly related to a change in blood flow "density" from the pulmonary hilum to the peripheral lung fields. A series of cardiac "density" changes and pulmonary hilar "density" changes are propagated directly after undergoing a kind of transformation. This is thought to occur with a slight phase difference from the relationship between cardiac "density" changes and pulmonary hilar "density" changes. Also, since changes in "density" in the pulmonary hilum and other areas are directly related to changes in blood flow "density" in the lung fields, it is possible to express the synchronization by reflecting it at the same rate (a relationship of agreement rate of Y≒X). Similarly, in the cervical vascular system and the major vascular systems of the chest, abdomen, pelvis, and limbs, it is thought that changes in "density" plotted in nearby central cardiovascular systems are directly related or related with a slight phase difference. Furthermore, this "density" fluctuates depending on the background, and when it propagates, the way in which the "density" changes is transmitted can be considered as a synchronization rate.

[0066] Here, for both the amount of change in a single image and the rate of change in a single image, we can assume that "total inspiratory volume ≈ total expiratory volume". Therefore, when calculating a relative value from the difference with respect to ambient air permeability, if we try to display it as a relative value (Standard Differential Signal Density / Intensity) with the amount of change set to 1 from the "density" of the lung field, we can depict the amount of change and rate of change for each of the following: (1) the difference image for each image, with each image set to 1 (normal assumption), (2) the total inspiratory volume or total expiratory volume with the "density (amount of change or rate of change)" added up from the difference image for each image, or the ratio with the absolute value of inspiratory and expiratory volume set to 1, and (3) the ratio with the total amount of "density" at each breath during multiple scans (selecting several times when it is 10%) set to 1.

[0067] Furthermore, in the case of 3D imaging such as MRI, the sum of the total "intensity" (in the case of MRI) and "density" (in the case of CT) of inspiration (with this value set to 1), and the difference between these "intensity" and "density," can be converted into the "peak flow volume data" of inspiration (even at rest or during labored breathing). By calculating the ratio of this "intensity" and "density," it is possible to convert the measured respiratory volume and respiratory rate in each part of the lung field, at least when calculating "3D × time" using MRI or CT. Similarly, by inputting the stroke volume, it is also possible to present an estimated value that converts the distribution of the "capillary phase" in the "flow" of the lung field into the distribution and volume of peripheral pulmonary blood flow.

[0068] In other words, (change in inspiratory volume per image) × (total number of inspiratory images) ≈ (change in expiratory volume per image) × (total number of expiratory images) ≈ (volume of inspiratory respiration at that time: natural or labored respiration) ≈ (volume of expiratory respiration at that time: natural or labored respiration) ≈ (change in inspiratory or expiratory volume of natural or labored respiration at that time). If we take only the change in one image, such as 10% or 20%, we can estimate the value by calculating (total number of images) × (change in volume over that time).

[0069] The extracted changes are visualized and depicted as images. This is the respiratory function analysis and vascular analysis described below. The rate of change of the thoracic cage and diaphragm is then visualized. At this time, artifacts to the results are removed again, and function is extracted from new data extraction waveforms, the initial base data waveforms, waveforms from other modalities, surrounding data, and multiple waveforms. The method for removing artifacts will be described later.

[0070] Furthermore, it is sometimes possible to grasp the characteristics even after excluding the altered components extracted from sources other than those mentioned above. For example, when understanding the movement of the abdominal intestines, the effects of respiration and blood vessels are excluded from the abdominal area to extract the movement of the abdominal intestines.

[0071] Embodiments of the present invention will be described below with reference to the drawings. Figure 1A is a diagram showing the schematic configuration of a diagnostic support system according to this embodiment. This diagnostic support system performs specific functions by having a computer execute a diagnostic support program. The basic module 1 consists of a respiratory function analysis unit 3, a pulmonary blood flow analysis unit 5, other blood flow analysis units 7, a Fourier analysis unit 9, a waveform analysis unit 10, and a visualization / digitization unit 11. The basic module 1 acquires image data from a database 15 via an input interface 13. The database 15 stores, for example, images in DICOM (Digital Imaging and Communication in Medicine). The image signal output from the basic module 1 is displayed on a display 19 via an output interface 17. Next, the functions of the basic module according to this embodiment will be described.

[0072] [Periodic analysis of respiratory elements] In this embodiment, the period of respiratory elements is analyzed based on the following indicators. "Respiratory elements" are a concept that includes all or part of exhalation or inhalation, as described above. That is, at least one frequency of respiratory elements is analyzed using at least one of the following: "density" / "intensity" in a certain region within the lung field, diaphragmatic movement, or thoracic movement. This "at least one frequency of respiratory elements" is a concept that includes cases where the frequency spectrum exhibited by the respiratory element is one or more and has a certain bandwidth. Considering the lung field as a collection of blocks, and since multiple frequencies are extracted from each block, in this embodiment these are processed as a frequency group. As mentioned above, the base data has both the concept of "wave morphology itself" and "wave interval (frequency: Hz)", so it is also possible to process it as wave morphology. Alternatively, data obtained from other measurement methods such as spirometry or external input information may be used, such as a range consisting of a certain volume "density" / "intensity" measured in a region with high X-ray (and other modalities such as CT and MRI) transparency.

[0073] Furthermore, by comparing the analysis results for each breath and analyzing trends from multiple data points, the accuracy of the data can be improved.

[0074] Furthermore, respiratory elements can be corrected by changing the phase of at least one frequency of the respiratory elements or by smoothing the waveform of the respiratory elements. In this case, the phase is matched to the wave using movements such as (thoracic cage, diaphragm movement) ≈ (thoracic cage movement) ≈ (density) ≈ (precise lung function) ≈ (thoracic cage sensor). In addition, the average "density" of the lung field is tracked, and the last change is approximated as the wave morphology using methods such as wave squaring to identify the wave. Here, in the case of "density" of the chest, the value that changes the most is the "density" of the lung, so the change in "density" of the lung may be evaluated by evaluating the "density" of the entire screen. When plotting waves, there may be a phase difference between the actual movement and the measured value. In that case, the phase may be corrected by the position of the maximum and minimum phase difference, or the overall wave morphology.

[0075] [Waveform analysis] From the waveform of the respiratory element, the frequency components of the waveform can be calculated. This allows for the acquisition of the "waveform synchronization image" described above. Specifically, the waveform of any part within the analysis range is identified, the frequency components of the identified waveform are extracted, and an image corresponding to the frequency components of the waveform is output.

[0076] [Cardiovascular analysis and vascular analysis] In this embodiment, cardiovascular rhythm analysis and vascular rhythm analysis are performed based on the following indicators. Specifically, the heart, pulmonary hilum, and major blood vessels are identified from measurement results of other modalities such as electrocardiograms and pulse meters, or from the lung contour, and the vascular rhythm is analyzed using changes in "density" and "intensity" of each area. Alternatively, the changes in "density" and "intensity" of the target area may be analyzed by manually plotting them on the image. Then, at least one frequency (waveform) of the cardiovascular rhythm element obtained from the heartbeat or vascular rhythm is identified. It is desirable to compare the analysis results for each beat and analyze trends from multiple data to improve the accuracy of the data. Furthermore, the extraction of "density" and "intensity" for each area can be performed multiple times or over a certain range to improve accuracy. Alternatively, the cardiovascular rhythm frequency or frequency band can be input.

[0077] [Lung field identification] Images are extracted from a database (DICOM), and the lung contour is automatically detected using the periodic analysis results of the respiratory elements described above. Conventional techniques can be used for this automatic detection of the lung contour. For example, the techniques disclosed in Japanese Patent Publication No. 63-240832 or Japanese Patent Publication No. 2-250180 can be used. Next, the lung field is divided into multiple block regions, and the changes in each block region are calculated. Here, the size of the block region may be determined according to the scanning speed. If the scanning speed is slow, it becomes difficult to identify the corresponding area in the next frame image of a given frame image, so the block region is made larger. On the other hand, if the scanning speed is fast, there are many frame images per unit time, so it is possible to track even if the block region is small. The size of the block region may also be calculated according to which timing in the period of the respiratory elements is selected. Here, it may be necessary to correct for the displacement of the lung field region. In that case, the movement of the rib cage, the movement of the diaphragm, the positional relationship of blood vessels throughout the lung field are identified, and the relative position of the lung contour is grasped and evaluated relatively based on that movement. Note that if the block area is too small, image flickering may occur. To prevent this, the block area needs to have a certain size.

[0078] The lung field can be represented as a coordinate system of points and control points using at least one Bézier curve in the automatically detected lung field region described above. Furthermore, it is possible to represent the lung field by using multiple closed curves, so-called "simple closed curves," enclosed by at least one Bézier curve. Similarly, it is possible to represent the object of analysis using one or more simple closed curves.

[0079] The lung fields in each frame can also be detected in other frames using at least one Bezier curve on the lung fields detected in a particular frame. For example, one method is to detect the largest and smallest lung fields and use these two lung fields to calculate the lung fields in the other frames. Here, a variable called "rate of change" is defined for the other frames. The "rate of change" is a value that represents the size of the lung fields, i.e., the state of respiration, and can be calculated from the position of the diaphragm or the average "intensity" of the entire image. It is also possible to calculate it using measurement data from external devices such as spirography, or to use a modeled rate of change. In this way, the variable "rate of change" can be arbitrarily defined, so for example, it can be calculated by assuming that the lung fields are changing at a constant rate (10%, 20%, 30%...). Since the rate of change defined in this way may contain errors, subsequent processing may be performed using the results of automatic or manual error removal, or the results of approximation using methods such as the least squares method. Assuming that the lung field is linearly transformed from its maximum to minimum size, the lung field in each frame is calculated using techniques such as linear transformation, based on the rate of change of each frame.

[0080] Furthermore, the above processing can be applied to any range of consecutive frames. For example, during respiration, the lung field repeatedly changes between maximum and minimum sizes, but in actual measurements, the shape at the maximum is not always constant. For example, by applying the above processing to each range from maximum to minimum, and from minimum to maximum, it is expected that the lung field can be calculated with higher accuracy than by defining and calculating two lung fields, the maximum and minimum. Here, we have used maximum and minimum as specific examples, but the present invention is not limited to this, and since it is an "arbitrary range," it is also possible to perform the processing at positions such as 0% and 30%, or 30% and 100% during respiration.

[0081] Furthermore, although accuracy is reduced, it is also possible to calculate the lung fields in each frame from a single lung field. For example, the change vector of the lung field can be defined by analogy with the shape of the rib cage, etc. Specifically, a method is adopted in which a change vector is defined for each control point of the Bézier curve, but the present invention is not limited to this. Then, the lung field in each frame is calculated using the detected lung field, the change vector, and the rate of change in each frame. The accuracy can be further improved by automatically or manually correcting this calculation result. The above method is also effective in 3D. That is, even in the case of 3D, it is possible to perform a process to detect lung fields in other frames using at least one or more Bézier surfaces on the lung field detected in a particular frame. This makes it possible to obtain images of lung fields between frames.

[0082] Figure 6C is a graph showing the cycle of respiratory elements. In the image of Figure 6C, a white vertical line is shown, which is a straight line (indicator) that shows the current position during the respiratory element cycle. It moves in accordance with the movement of the lungs shown in the video of the lungs in Figure 6B to show the current position during the respiratory element cycle. By showing the current position during the respiratory element cycle, it becomes possible to clearly understand the current position in the cycle of lung movement. Furthermore, in this invention, not only is the cycle of respiratory elements represented graphically, but it is also possible to graph everything that is linked to lung movement, such as blood flow "density," the rib cage, and the diaphragm.

[0083] Furthermore, if the subject "holds their breath," it may not be possible to identify the frequencies of the respiratory elements. In this case, Fourier analysis, as described later, is performed using at least one frequency of the cardiovascular elements extracted from the subject's heartbeat or vascular beat. In this case, the method of dividing the block region, as described later, may be changed depending on the movement of the heart, diaphragm, or other dynamic parts linked to respiration.

[0084] [Edge detection and evaluation] This invention makes it possible to detect and evaluate the edges of the lung. For example, after calculating the lung field using the method described above, the position and shape of the edges can be detected again with high accuracy. Points are plotted at arbitrary positions within the calculated lung field, lines are extended in all directions from there, and the change in pixel values ​​is evaluated along each line. For example, as shown in Figure 14, when pixel values ​​are calculated along a line segment S that cuts the lung, it can be seen that pixels fluctuate greatly at the edges, but the absolute values ​​of these fluctuations are different. For example, by adjusting the thresholds for detecting the left edge and the right edge, the accuracy of edge detection can be improved. It is also possible to utilize the characteristics of pixel value fluctuations for each region. As shown in Figure 14, even if the difference between the edges of regions S2 and S3 is small, the edges of regions S2 and S3 can be identified from the variance of the pixel value fluctuations. Here, we have focused on variance, but this invention is not limited to this.

[0085] Furthermore, a similar approach makes it possible to detect the boundaries of the analysis range for organs other than the lungs, blood vessels, tumors, etc. For example, when contrast agent is present in a blood vessel, the inside of the blood vessel can be clearly visualized, but it is not easy to accurately calculate the outside of the blood vessel or its thickness. In this embodiment, since the boundaries can be accurately detected, the shape and characteristics of the blood vessels within the analysis range can be calculated. This makes it possible to quantitatively understand the thickness and circumference of blood vessels, which were previously difficult to grasp, and use this information for diagnosis.

[0086] [Create a block region] This section describes methods for dividing the lung field into multiple block regions. Figure 1B shows a method for dividing the lung field radially from the hilum. Since the lung moves more towards the diaphragm than towards the apex, it may be advisable to plot points that are more coarsely divided closer to the diaphragm. In Figure 1B, vertical lines (dotted lines) may be added to divide the lung into multiple rectangular (square) block regions. This allows for a more accurate representation of lung movement. It is also possible to divide the lung field using methods such as: "plotting points vertically across the lung to divide it crosswise," "plotting points horizontally across the lung to divide it longitudinally," "drawing tangents at the lung apex and the diaphragm, defining the intersection point as the center point, and dividing the lung with line segments drawn at a certain angle from a line (e.g., a vertical line) containing that point," and "cutting the lung with multiple planes perpendicular to a line connecting the lung apex (or hilum) to the end of the diaphragm." These methods are also applicable to three-dimensional stereoscopic images. In 3D models, each organ is perceived as a space enclosed by multiple curved or flat surfaces. Organs can also be further subdivided. For example, a 3D model of the right lung can be divided into the upper, middle, and lower lobes.

[0087] The lung field region should be evaluated relatively based on its movement, by identifying the movement of the rib cage, the movement of the diaphragm, and the positional relationship of blood vessels throughout the lung field, and by understanding the relative position of the lung contour. For this reason, in the present invention, after automatically detecting the lung contour, the region identified by the lung contour is divided into multiple block regions, and the values ​​of image change (pixel values) included in each block region are averaged. For example, as shown in Figure 10, it is also possible to plot points on the edges of opposing lungs on a Bézier curve, connect them, and then use a curve passing through the intermediate point. As a result, as shown in Figure 1C, even if the morphology of the lung changes over time, it becomes possible to track the temporal changes of the region of interest. On the other hand, Figure 1D is a diagram showing the temporal changes when the organ to be analyzed (in this case, the lung) is divided into block regions without considering its morphology. As mentioned above, the lung field region should be defined by identifying the movement of the rib cage, the movement of the diaphragm, and the positional relationship of blood vessels throughout the lung field, thereby understanding the relative position of the lung contour and evaluating it relatively based on that movement. However, as shown in Figure 1D, if the area is divided into block regions without specifying the lung field region, the area of ​​interest may move outside the lung field region due to changes in the lung over time, resulting in a meaningless image. In particular, since the movement of the diaphragm has a strong effect in contracting the lung field, it is preferable to correct the lung field region by incorporating rib cage components and other multiple elements, rather than just correcting the diaphragm or overall values. Alternatively, respiratory element frequencies or frequency bands can be input. The same region division calculations can be performed for 3D as well.

[0088] Furthermore, as shown in Figure 11, in lung field A, it is also possible to use a Bézier curve to select internal control points within the detected lung field and divide the lung field with curves or straight lines passing through these internal control points. That is, control points are set not only within the framework of the lung field but also within the lung field region, and these control points are used to divide the lung field region (A). In this case, as shown in Figure 12, the spacing of control points at the outer edge of the detected lung field and in its vicinity may be made relatively large (1), and the spacing of internal control points may be made relatively small (2) according to the expansion ratio of each part within the detected lung field. In addition, in lung field A, the spacing of control points may be made relatively large as they move in the head-to-tail direction relative to the human body, or relatively large according to a specific vector direction. The method of determining this vector is arbitrary, but for example, it may be set in the direction from the lung apex toward the opposite side of the lung field, or as shown in Figure 1B, it may be set in the direction from the lung hilum toward the opposite side of the lung field. It is also possible to set the vector in a direction corresponding to the structure of the lung. In this way, by dividing the lung field into "unequal divisions," it becomes possible to display images that take into account the characteristics of each region. For example, the outer periphery of the lung field has large movements and large displacements, so the blocks are made larger, while the interior of the lung field has small movements and small displacements, so the blocks are made smaller and more detailed. Also, for example, the diaphragmatic side of the lung field has large movements and large displacements, so the blocks are made larger, while the head side of the lung field has small movements and small displacements, so the blocks are made smaller and more detailed. This makes it possible to improve the accuracy of the display. This method is not limited to the lung field, but can also be applied to dynamic areas that are linked to respiration. This method can also be applied when dividing the lung three-dimensionally into lobes. It can also be used when enclosing the area below the diaphragm, for example, the heart or other organs, with Bézier curves. In this case as well, it is possible to define vectors in a direction according to the structure of the heart or other organs and divide the region unequally.

[0089] Next, the image data is interpolated to eliminate artifacts. That is, if bones or other objects are included in the analysis range, they will appear as noise, so it is desirable to remove the noise using a noise-cutting filter. In X-ray images, air is typically set to -1000 and bone to 1000, so areas with high transmittance have low pixel values ​​and are displayed in black, while areas with low transmittance have high pixel values ​​and are displayed in white. For example, if the pixel value is represented by 256 levels, black is 0 and white is 255. In the lung field, areas around locations where there are no blood vessels or bones are easily penetrated by X-rays, so the pixel values ​​of the X-ray image are low and the X-ray image appears black. On the other hand, areas where blood vessels or bones are present are difficult for X-rays to penetrate, so the pixel values ​​of the X-ray image are high and the X-ray image appears white. The same can be said for CT and MRI. Here, from the results of the periodic analysis of the respiratory elements described above, it is possible to interpolate the data using values ​​of the same phase based on the waveform per breath and eliminate artifacts. Furthermore, if "different coordinates," "extreme fluctuations in pixel values," or "abnormally high frequency or density" are detected, cutoffs can be applied to these, and the remaining obtained images can be used for diaphragm Hz calculation and lung field adjustment by identifying a continuous and smooth wave shape using, for example, the least squares method. When superimposing images, there are two methods: (1) acquiring one image before and one after and superimposing the acquired comparison image while maintaining its coordinates, and (2) acquiring one image before and one after as a base, then relatively expanding the image and superimposing its relative position information onto the base. Through these methods, it becomes possible to correct the morphology of the lung fields or correct changes in block region images. At that time, artifacts in the results are again excluded, and functions are extracted from the new data extraction waveform, the initial base data waveform, waveforms of other modalities, surrounding waveforms, and multiple waveforms. The number of times this is done can be one or multiple times.

[0090] Here, we will explain "reconstruction" in terms of the time axis. For example, if the intake time is 2 seconds at 15 f / s, 30+1 images will be obtained. In this case, 10% "reconstruction" can be performed simply by stacking 3 images together. However, if, for example, 0.1 seconds is 10%, and only images taken at 0.07 seconds and 0.12 seconds are available, then "reconstruction" of 0.1 seconds is necessary. In this case, the "reconstruction" is performed by using an intermediate value (the average of both) of the images around 10%. Alternatively, the coefficient can be changed according to the proportion of time. For example, if there is a difference in the time axis and there is no image taken at 0.1 seconds, but there are images taken at 0.07 seconds and 0.12 seconds, the "reconstruction" can be performed by recalculating as "(the value for 0.07 seconds) × 2 / 5 + (the value for 0.12 seconds) × 3 / 5". Furthermore, the system recognizes the positional relationship of changes within a given second from the average respiration and the change in the diaphragm coefficient, and uses these values ​​as coefficients to determine the percentage of the numbers. It is desirable to calculate with thickness, including 0-100% of "Maximum Differential Intensity Projection," such as 10-20% "reconstruction" or 10-40% "reconstruction." In this way, it is possible to perform "reconstruction" at a rate of one respiration even for parts that have not been filmed. Moreover, this invention can perform "reconstruction" not only on respiration, but also on blood flow, chest wall movement, diaphragm, and a series of other movements linked to these. It is also possible to perform "reconstruction" block by block or pixel by pixel. It is desirable to calculate with thickness, including 0-100% of "Maximum Differential Intensity Projection," such as 10-20% "reconstruction" or 10-40% "reconstruction."

[0091] Alternatively, the lung fields can be detected using the methods described above, and these detected lung fields can be normalized. That is, the detected lung fields can be spatially normalized, or temporally normalized using reconstruction. Although the size and shape of lung fields differ due to individual differences, normalization allows them to be displayed within a certain area.

[0092] [Diaphragm and rib cage] By identifying the lung fields as described above, it becomes possible to understand the movement of the diaphragm and the rib cage. Specifically, by calculating the curves of the diaphragm and the rib cage on the X-ray (2D image) of the recognized diaphragm as a collection of fine coordinates, and quantifying the average, the rate of downward change and the amount of change in the local part of the curve, and the deformation rate by "curve fitting" the diaphragm as a curve, the functional evaluation from the image can be positioned. Similarly, the curves of the edges drawn in the chest other than the diaphragm plane can also be calculated as a collection of fine coordinates, and functional evaluation from the image can be performed by quantifying the average and the rate of change of the curve. The above two rates of change and changes are evaluated as relative and interconnected, and functional evaluation of movement is performed by quantifying and imaging areas with different rates of change (such as areas that do not move in the same interconnected way).

[0093] Here, we will explain the "Diaphragm and Thoracic Cage Evaluation Method." First, for the diaphragm, its movement is displayed using left and right horizontal lines perpendicular to the body's axis (the so-called midline) as the axis. Next, the diaphragm line is flattened with respect to the baseline. That is, the diaphragm line is aligned with a horizontal straight line. Then, the movement of the diaphragm is evaluated. Furthermore, the diaphragm line is stretched and flattened, and the movement of the curves perpendicular to each other is evaluated. Next, on the lateral side of the thoracic cage, the movement is evaluated using the line connecting the lung apex to the diaphragm-thoracic angle as the baseline (axis). The thoracic cage line is flattened with respect to the baseline, that is, the thoracic cage line is aligned with the straight line of the "lung apex-costopharynx angle," and the movement is evaluated. The thoracic cage line is stretched and flattened with respect to the baseline, and the movement of the curves perpendicular to each other is evaluated. Then, the curvature and radius of curvature of the above thoracic cage and diaphragm lines are evaluated. Finally, the above changes are calculated as the "amount of change," and this amount of change is differentiated to evaluate it as the "rate of change."

[0094] Figures 6B and 6C show examples of images displayed on a screen. In Figure 6B, the movement of the left lung is displayed as a video. In the image of Figure 6B, a white horizontal line is shown; this is a straight line (indicator) indicating the position of the diaphragm, and when the video is played, it moves up and down following the movement of the diaphragm. In this way, by detecting the diaphragm and showing an indicator of the detected diaphragm's position, i.e., the white horizontal line indicating the diaphragm's position, physicians can perform image diagnosis. Furthermore, by using the recognition of the lung field-diaphragm line, not just a part of the diaphragm, it becomes possible to recognize all points and diagnose a region of the diaphragm such as left and right, or lateral and medial, or even the entire diaphragm. Similarly, not only the diaphragm, but also the movement of dynamic parts linked to respiration, such as the rib cage, can be similarly determined by straight lines such as tangential positions or straight lines of the rib cage recognized by lung field recognition. In this way, assuming that the edges are moving, it becomes possible to detect the edges by taking the difference between consecutive images. For example, tumors are often hard, while the surrounding area is soft. Therefore, since the tumor itself does not move much, while the surrounding area moves actively, the tumor's edge can be detected by taking the difference.

[0095] Furthermore, in 3D images such as MRI and CT, the diaphragm surface can be treated as a single coordinate or a three-dimensional curved surface. By calculating this coordinate or curved surface as a collection of fine coordinates (the contour of the diaphragm's edge, a plane, and a set of coordinates), and quantifying the average, the rate of downward change or the amount of change in a localized area of ​​the curved surface, and the deformation rate by "curve fitting" the diaphragm as a curved surface, the functional evaluation from the image can be positioned. Similarly, the curved surfaces of the edges drawn in the chest region other than the diaphragm surface can also be calculated as a collection of fine coordinates, and by quantifying the average and the rate of change of the curved surface, functional evaluation from the image can be performed. The above two rates of change and changes are evaluated relatively and in conjunction with each other, and by quantifying and visualizing areas with different rates of change (such as areas that do not move in the same way), functional evaluation of movement can be performed.

[0096] [Fourier analysis] Based on the respiratory element period and vascular rhythm period analyzed as described above, Fourier analysis is performed on the "density" / "intensity" values ​​and their changes in each block region. Figure 2A shows the "intensity" change of a specific block and the result of its Fourier analysis. Figure 2B shows the Fourier transform result with frequency components close to the heartbeat extracted, and the inverse Fourier transform of this result showing the "intensity" change of frequency components close to the heartbeat. For example, performing a Fourier transform (Fourier analysis) on the "intensity" change of a specific block yields the result shown in Figure 2A. Then, extracting the frequency components close to the heartbeat from the frequency components shown in Figure 2A yields the result shown on the right side of Figure 2B. By performing an inverse Fourier transform on this, a "intensity" change synchronized with the heartbeat can be obtained, as shown on the left side of Figure 2B.

[0097] As shown in Figure 9, it is also possible to weight specific spectra by multiplying them by coefficients. For example, this method can be used to achieve waveform tuning. That is, when performing the inverse Fourier transform, multiple frequencies can be selected, multiplied by their ratio, and then the inverse Fourier transform can be performed. For example, if you want to highlight the spectrum with the highest frequency within the extracted band, you can double its spectral intensity. In this case, frequency continuity is not required. It is possible to select spectra that exist discretely.

[0098] Furthermore, the location of the heart's "density" can be inferred from the morphology of the left lung (or the right lung in cases of situs inversus, etc.) (the area of ​​the indented part of the left lung from the morphology of the lung field extraction), as well as the positions of the vertebrae and diaphragm. In this case, the heart's ROI is taken to extract the "density." When performing this extraction, a rough area is used to infer from the relative spectral values ​​of respiration and blood flow. In addition, "filtering" may be performed beforehand using the Hz band generated by cardiovascular beats (heart rate 40-150 Hz, ≈ 0.67 Hz-2.5 Hz) to remove frequencies caused by respiration and other "artifacts." Also, since the position of the heart changes according to the respiratory situation, the relative position of the heart may be changed from the morphological values ​​of the rib cage as the position of the rib cage changes, allowing for more accurate extraction of cardiovascular beats, the pulmonary hilum, and major vessels. Furthermore, there is a method to calculate frequencies based on the regularly moving contour of the heart, similar to the movement of the diaphragm.

[0099] Here, when performing an inverse Fourier transform on a spectrum consisting of frequency components, both frequency elements identified from the "density" of respiration and blood flow (respiratory frequency, cardiovascular beat frequency) and the bandwidth of the spectrum (a band-pass filter (BPF) may be used) may be taken into consideration, or the inverse Fourier transform may be performed based on either one of these elements. Furthermore, from the spectrum obtained after the Fourier transform, at least one frequency may be selected for the inverse Fourier transform based on the spectral composition ratio of organ-specific periodic changes. In addition, it is possible to identify the waveform of a specific organ or region to be analyzed based on the composition ratio of multiple frequencies obtained after the Fourier transform (creation of waveform-synchronized images).

[0100] Furthermore, when performing a Fourier transform, the Autoregressive Moving Average (AR) method can be used to enable faster calculations. The AR method employs the Yule-Walker equation and Kalman filter in an autoregressive moving average model, and the Yule-Walker estimates derived from these, along with the PARCOR method and the least squares method, can be used to supplement the calculations. This makes it possible to acquire images more quickly, in near real-time, and to assist in calculations and correct for artifacts. Through such Fourier analysis, it becomes possible to extract and display the image properties of each block region.

[0101] Furthermore, it is possible to employ a "digital filter" method during this Fourier analysis. That is, a "digital filter" is used, which involves performing a Fourier transform on the original waveform to obtain the parameters of each spectrum, and then performing calculations on the original wave. In this case, a digital filter is used instead of performing an inverse Fourier transform.

[0102] Here, the image changes of each block region in each frame image are Fourier transformed, and from the spectrum obtained after the Fourier transform, a spectrum within a certain band that includes the spectrum corresponding to the period of the respiratory element can be extracted. Figure 2C shows an example of extracting a certain band from the spectrum obtained after the Fourier transform. The frequency f of the composite wave spectrum has the relationship "1 / f = 1 / f1 + 1 / f2" between it and the original frequencies f1 (respiratory component) and f2 (blood flow component), and the following method can be used when extracting the spectrum.

[0103] (1) Extract the portion with a high spectral ratio of blood flow. (2) The spectrum is extracted by dividing it midway between the peak of the spectrum corresponding to respiration / blood flow and the peaks of multiple composite waves in its vicinity. (3) The spectrum is extracted by dividing it into the peak corresponding to respiration / blood flow and the trough of the combined wave spectrum in its vicinity.

[0104] As described above, the present invention does not use a fixed BPF, but rather extracts spectra within a certain bandwidth that include spectra corresponding to the period of respiratory elements. Furthermore, in the present invention, it is also possible to extract spectra within a certain bandwidth from the spectra obtained after the Fourier transform that include frequencies other than respiratory elements obtained from frame images (for example, "density" / "intensity" of each part, heart rate elements obtained from heartbeats or vascular beats), or spectra corresponding to frequencies input externally by the operator (for example, spectral models).

[0105] Here, if the composite wave spectrum consists of only two components (respiration and blood flow), the distribution will be 50% + 50%, and if there are three components, the distribution will be 1 / 3 each. Therefore, the composite wave spectrum can be calculated to some extent from the spectral components and their heights, such as what percentage of the spectrum is for the respiratory component and what percentage is for the blood flow component. It is possible to extract the spectrum where that ratio (%) is high. That is, the ratio of the blood flow component / respiratory component to the composite wave component is calculated, and the spectral value of the blood flow component / respiratory component with a high value is calculated and extracted. In addition, in the identification of the diaphragm, etc., it is sometimes the case that only the spectrum or superposition of spectra corresponding to the region where the Hz (frequency) is relatively constant, i.e., the region where the Hz change is small, is extracted from the data obtained from the frequencies of respiration and cardiovascular systems. Furthermore, when determining the spectral bandwidth, in the case of diaphragm identification, etc., the spectral bandwidth may be determined in the range where the Hz change occurs and the surrounding region. The constituent elements of the waveform may also be taken into consideration.

[0106] Furthermore, when performing an inverse Fourier transform, it is possible to choose between two approaches for the spectrum: "extracting from a simplified model of frequencies and frequency bands using one or more high-frequency regions (simulation approach)" and "extracting high-frequency or low-frequency regions based on actual frequencies or frequency bands, according to the spectral values ​​(field approach)." Also, if the frequency of the heart is A and the frequency of the lungs is B, B can be obtained by subtracting A from the entire frequency band. In addition, it is possible to extract not just one point but multiple points on the frequency axis from the spectrum obtained from the Fourier transform.

[0107] As a result, it is possible to extract spectra that not only perfectly match the respiratory element period and vascular period, but also those that should be considered, thereby contributing to diagnostic imaging. It is known that "respiration" and "heartbeat" fall within specific frequency bands. Therefore, for example, in the case of respiration, it is possible to pre-specify respiratory frequencies and circulatory frequencies using filters such as "0~0.5Hz (respiratory rate 0~30 breaths / min)" and in the case of the circulatory system, for example, "0.6~2.5Hz (heart rate / pulse rate 36~150 beats / min)". This makes it possible to display frequency-synchronized images. This is because when acquiring cardiac "density" changes, respiratory (lung) "density" changes may be picked up, and vice versa.

[0108] [Visualization and Quantification] The results of the analysis described above are visualized and quantified. In this specification, a "modeled lung" is defined when visualizing and quantifying the results. When displaying the lungs as moving images, relative judgment is difficult because the positional relationships change. Therefore, the discrepancies in positional relationships are spatially unified and averaged. For example, the shape of the lungs is fitted to a figure such as a sector and displayed in a symmetrical form. Then, the concept of reconstruction is used to unify it temporally. For example, the "state of 20% of the lungs from multiple breaths" can be extracted and defined as the "state of 20% of the lungs in one breath." In this way, a lung that has been spatially and temporally unified is called a "modeled lung." This makes it easier to make relative judgments when comparing different patients or comparing the present and past of a single patient.

[0109] For example, as a standard uptake, the "density" and "intensity" of the entire lung field measured may be displayed relatively / logarithmically with the average value set to 1. Also, by only considering the direction of blood flow, changes in a specific direction may be isolated. This makes it possible to extract only meaningful method data. Using the lung field identification results, pseudo-colorization is performed to track changes in the analysis range. That is, the analysis results of each individual (subject) are fitted to a relative region according to a specific shape (minimum, maximum, mean, median) that matches the phase.

[0110] Furthermore, the model is transformed into a specific shape and phase that allows for comparison of multiple analysis results. Additionally, when creating the modeled lung, the relative positional relationships within the lung field are calculated using the results of the periodic analysis of the respiratory elements mentioned above. The modeled lung is created using a line that comprehensively averages the thoracic lines, density, and diaphragm of multiple patients. When creating the modeled lung, for pulmonary blood flow, the distance can be measured radially from the hilum to the lung apex. For respiration, corrections are necessary according to the movement of the thoracic wall and diaphragm. Furthermore, a complex calculation considering the distance from the lung apex may also be performed.

[0111] Furthermore, after the inverse Fourier transform, only the blocks with relatively large amplitude values ​​can be extracted and displayed. That is, when performing Fourier analysis on each block, after the inverse Fourier transform, there will be blocks with large wave amplitudes and blocks with small wave amplitudes. Therefore, it is also effective to extract and visualize only the blocks with relatively large amplitudes. In addition, after the inverse Fourier transform, the real and imaginary parts of each value can be used separately. For example, it is possible to reconstruct an image from only the real part, from only the imaginary part, or from the absolute values ​​of both the real and imaginary parts.

[0112] Fourier analysis can be performed on a modeled lung. That is, the modeled lung can be used when combining images of respiratory rate, performing Fourier analysis, or determining relative position. By using a modeled lung, it is possible to apply multiple acquired frames to the modeled lung, or in the case of blood vessels, to a modeled lung calculated according to the heartbeat (for example, the heartbeat obtained from the hilum of the lung), thereby making the relative position constant when performing Fourier analysis. By using a modeled lung when acquiring the base respiratory state, it is also possible to obtain stable calculation results. Furthermore, by modeling the lung, spatial differences can be fixed, making it possible to visualize lung movement more easily.

[0113] In image processing, the relative evaluation labeling method is as follows: The image is labeled relatively using black and white or color mapping. A few percent of the "density" / "intensity" values ​​obtained by the difference are sometimes cut, and the remaining values ​​above and below are displayed relatively. Alternatively, since the values ​​around a few percent before and after the obtained difference may be exceptionally high, these may be removed as "artifacts," and the remaining parts may be displayed relatively. In addition to methods such as 0-255 gradation, values ​​may also be displayed as 0-100%.

[0114] It is also possible to display pixels somewhat ambiguously, blurring the overall image. In particular, in the case of pulmonary blood vessels, low signal values ​​are mixed in with high signal values, but if only the high signal values ​​can be roughly grasped, the overall ambiguity is acceptable. For example, in the case of blood flow, signals above the threshold can be extracted, while in the case of respiration, signals above the threshold can not be extracted. Specifically, if the numbers in the following table are treated as one pixel and the middle value is obtained, the proportion occupied by the middle value can be obtained and averaged within one pixel, allowing for a smooth representation between adjacent pixels. This method can also be used when calculating the average intensity for each block. [Table 1]

[0115] This method can be applied not only to detecting density in the lung field but also to detecting density in any analysis range and removing areas where density changes significantly. It also cuts off points that significantly exceed a pre-set threshold. Furthermore, it recognizes and removes rib morphology, such as suddenly appearing high / low signal lines. Similarly, it may remove sudden signals that appear suddenly from the phase, such as the characteristic of patients where artifacts are observed around 15%-20% of the reconstruction phase, which differ from normal wave changes. Note that when initially acquiring base data, the phase may differ in calculations such as (diaphragm) ≈ (thoracic cage) ≈ (thoracic cage movement) ≈ (spirometer) ≈ (lung field), and (density) ≈ (volumetry) of the field, and in such cases, the phase may be applied to the morphology that can actually be recognized (XP contour).

[0116] Once a modeled lung is created, as described above, it becomes possible to quantify and present the synchronicity, agreement rate, and mismatch rate (display of frequency-synchronized images or wavelength-synchronized images). This allows deviations from a normal state to be displayed. According to this embodiment, by performing Fourier analysis, it becomes possible to discover the possibility of new diseases, compare with a normal person, compare hands and feet, and compare with the opposite hand and foot. Furthermore, it becomes possible to understand what is abnormal in foot movement, swallowing, etc., by quantifying the synchronicity. It also becomes possible to determine whether a person with a disease has changed after a certain period of time, and if so, to compare before and after the change. In addition, by making the lung field a shape that is easy to view radially (circular to nearly oval) with a constant distance from the periphery, it becomes easier to evaluate the inner to middle layer, outer layer, etc., and it is also possible to represent whether "peripheral dominance of blood vessels" or "middle layer dominance" is possible.

[0117] Furthermore, during visualization, it is possible to switch between displaying the image after the Fourier transform and the image before the Fourier transform, or to display both side by side on a single screen.

[0118] As shown in Figure 2D, when the modeled lung is set to 100, it is possible to determine the percentage difference in the human body and display the rate of change. Furthermore, it is possible to determine the difference not only for the entire lung but also for a part of the lung. In particular, as mentioned above, it is possible to identify only the movement of the diaphragm, fix the shape of the lung fields other than the diaphragm, and display the movement of the diaphragm, as well as the synchronization rate and rate of change. It is also possible to fix the entire lung field and display the synchronization rate and rate of change. In addition, by performing "Variation classification," it is possible to identify standard blood flow. That is, it is possible to identify the period of respiratory elements, calculate the relative positional relationship of blood vessels, and identify the blood flow dynamics of the subject as standard blood flow.

[0119] Alternatively, lung detection can be performed using pattern matching techniques. Figures 2E to 2H show examples of pattern images of lung fields. As shown in Figures 2E to 2H, the shape of the lungs can be classified into patterns, and the closest matching patterns can be extracted. This method makes it possible to determine whether the target image represents one lung or both lungs. It can also determine whether it is the right lung or the left lung. There is no limit to the number of patterns, but it is expected that there will be 4 to 5 patterns. In addition, there is a method to recognize the right lung, left lung, or both lungs based solely on the morphology (shape) of the lung field. Furthermore, it is also possible to recognize a thick band-shaped "area of ​​decreased transparency" formed by the vertebral bodies and mediastinum, and to recognize the left or right or both lungs based on the positional relationship between this band-shaped area of ​​decreased transparency and the "area of ​​increased transparency" in the lung field. As shown in Figure 2H, this method can also be applied to the area below the diaphragm. This makes it possible to recognize the area below the diaphragm and the heart.

[0120] Furthermore, since air has the highest permeability and is even more permeable than the lung fields, it is desirable to take air into consideration in the calculations. In other words, the position of air on the screen can be used to make the following determinations. If the area of ​​air in the upper right corner of the screen is greater than the area of ​​air in the upper left corner of the screen, it will be recognized as the left lung. This is because the area of ​​air outside the human body is larger in the shoulder area during imaging. If the area of ​​air in the upper left of the screen is greater than the area of ​​air in the upper right of the screen, it is recognized as the right lung. This is because, as mentioned above, the area of ​​air outside the human body is larger in the shoulder area during imaging. Next, if the area of ​​air in the upper right corner of the screen is approximately equal to the area of ​​air in the upper left corner of the screen, it will be recognized as both lungs. This is because the areas of air are roughly the same size on both sides.

[0121] Furthermore, air from the intestines can enter the area beneath the diaphragm, which may cause it to be obscured. Therefore, it is possible to first identify the general areas of reduced transparency within and around the lung field, starting from the center of the lung, and then identifying the edges of the lung field along these lines. This technique is, for example, It is also possible to use the techniques disclosed at "https: / / jp.mathworks.com / help / images / examples / block-processing-large-images_ja_JP.html".

[0122] This allows for comparison and quantification between one patient and another. It also enables comparison and quantification between normal lungs or blood vessels and typical abnormal lung function or blood flow. Furthermore, modeled lungs and standard blood flow can be used to relatively evaluate lung function and pulmonary blood flow in a patient at different time points. Such modeled lungs and standard blood flow can be compiled from typical examples of various types of patients and healthy individuals, and used as indicators when morphologically applying them to a particular patient for evaluation.

[0123] [Drawing of the lung field] Generally, the lung field contains ribs, which have low transparency, making it difficult to mechanically identify the lung contour using only "density" as an indicator. Therefore, this specification employs a method of provisionally drawing the lung field contour using a combination of Bézier curves and straight lines, and then adjusting the lung contour to improve the accuracy of the match.

[0124] For example, if the contour of the left lung is represented by four Bézier curves and one straight line, it becomes possible to draw the lung contour by finding five points on the contour and four control points. By shifting the position of the points and drawing multiple lung contours, and evaluating the matching using conditions such as "maximizing the sum of 'density' within the contour" or "maximizing the difference between the sum of 'density' of a few pixels inside and outside the contour line," it becomes possible to detect the lung contour with high accuracy. In practice, it is also possible to identify the positions of several points from the contour of the upper part of the lung, where edges are relatively easy to detect, or from the position of the diaphragm detected using the method described later, thereby reducing the number of simulation trials mentioned above. It is also possible to extract points close to the outer edge using classical binarization for contour extraction and adjust the control point positions of the Bézier curve using the least squares method, etc.

[0125] Figures 3A and 3B show examples of drawing the contour of the lung field using both Bézier curves and straight lines. Figure 3A shows the case where the lung area is at its maximum (maximum contour), and Figure 3B shows the case where the lung area is at its minimum (minimum contour). In each figure, "cp1~cp5" indicates control points, and "p1~p5" indicates points on the Bézier curve or straight line. In this way, once the maximum and minimum contours are understood, it becomes possible to calculate the intermediate contours. For example, it becomes possible to display the states of 10%, 20%, etc., of exhaled air. Thus, according to this embodiment, it is possible to draw at least the lung field, blood vessels, or heart using at least one Bézier curve. Note that the above method is not limited to the lungs, but can be applied to other organs as "organ detection". Furthermore, it is possible to use at least one Bezier curve on a predetermined analysis range (such as the boundary between a tumor, the hypothalamus, basal ganglia, and internal capsule of the brain) in a specific frame, and then perform a process to detect the corresponding range in other frames.

[0126] Furthermore, this method is applicable not only to planar images but also to three-dimensional images (3D images). By defining the equation of a curved surface and setting its control points, it becomes possible to analyze a region enclosed by multiple curved surfaces.

[0127] [Detection of movement of the diaphragm or other dynamic parts associated with respiration] In a series of images, it is possible to detect the movement of the diaphragm or other dynamic parts associated with respiration. When images are selected at arbitrary intervals from a series of images and the difference between them is calculated, the difference will be particularly large in areas with high contrast. By appropriately visualizing this difference, areas where movement has occurred can be detected. During visualization, noise reduction using thresholds or curve fitting using methods such as the least squares method can be used to emphasize the continuity of areas with large absolute differences.

[0128] In the lung field, the contrast of the lines where the diaphragm and heart meet is prominent. As shown in Figure 4A, by taking the difference between two lung images and visualizing the difference with a certain threshold, the lines where the diaphragm and heart meet can be visualized, as shown in Figure 4B.

[0129] [Estimation of diaphragm movement] In this method, while the diaphragm's position can be detected when it is moving between target images, it is difficult to detect areas where the diaphragm's movement slows down. In other words, detection is difficult at the timing of the transition between exhalation and inhalation, while breathing is held, and immediately after the start or end of imaging. In this method, the movement of the diaphragm is estimated using an arbitrary interpolation method.

[0130] Using the method described above, the diaphragm line was visualized as shown in Figure 4B. Then, the 1024px vertical image was divided into 128 rectangles of 8px each, the signal values ​​contained in each rectangular area were summed, and a bar graph was generated as shown in Figure 4C. Among the multiple peaks, the peak at the lowest coordinate, indicated by the dotted rectangle, is expected to represent the position of the diaphragm. In a normal standing XP image, the diaphragm is displayed as a curve, and this coordinate is approximated as the position of the diaphragm.

[0131] When the diaphragm position was detected for all images using this method, a "peak position" was detected as shown in Figure 5. By correcting this detected value, the movement of the diaphragm was estimated. First, if the difference was greater than a certain value, it was considered an outlier and excluded (thin solid line in Figure 5). The data from which the outliers had been excluded was divided into arbitrary clusters, and a quartic curve regression was performed on each cluster, and the results were joined together (thick solid line in Figure 5). Regression analysis was performed in this analysis, but the present invention is not limited to this, and any interpolation method such as spline interpolation can be used.

[0132] [Refining dynamic site detection] The contrast of dynamic areas may not be uniform along the lines. In such cases, the shape of dynamic areas can be detected more accurately by changing the threshold used for noise reduction and performing the detection process multiple times. For example, in the left lung, the contrast of the diaphragm line tends to weaken as it moves towards the interior of the body. In Figure 4B, only the right half of the diaphragm is detected. In this case, by changing the threshold setting used for noise reduction, the remaining left half of the diaphragm can also be detected. By repeating this process multiple times, it becomes possible to detect the shape of the entire diaphragm. This method makes it possible to quantify not only the position of the diaphragm but also the rate and amount of change of lines and surfaces regarding its shape, which can be useful for new diagnoses.

[0133] The detected position or shape of the diaphragm can be used for diagnosis. Specifically, in this invention, the coordinates of the diaphragm are graphed, and the coordinates of the rib cage and diaphragm are calculated using the curves (surfaces) or straight lines calculated as described above. Furthermore, heart rate, vascular rate, and lung field "density" can be graphed as positions and coordinates corresponding to their cycles. Such methods can also be applied to dynamic parts that are linked to respiration.

[0134] This method allows for measurement not only of Hz during inspiration and expiration, but also of changes in the frequency (Hz) of the diaphragm or other dynamic parts linked to respiration, in a frequency band corresponding to those changes. Furthermore, during spectral extraction of the band-pass filter (BPF), it becomes possible to set the BPF within a certain range according to the state of each respiration, and to create a variable BPF where the axis of the BPF position changes during the "reconstruction phase" of each respiration, potentially creating an optimal state. This makes it possible to provide images that correspond to fluctuations in the respiration rhythm, such as when breathing slows down or stops (Hz=0).

[0135] Furthermore, the overall frequency of exhalation or inhalation may be calculated based on the proportion of the respiratory element to the total exhalation or inhalation. In addition, when detecting the diaphragm, multiple measurements may be performed, and the one with the most stable signal or waveform may be selected. As a result, it becomes possible to calculate the frequency of at least one respiratory element from the detected position or shape of the diaphragm, or the position or shape of a dynamic part linked to respiration. Once the position or shape of the diaphragm or dynamic part is known, the frequency of the respiratory element can be determined. This method allows tracking of the subsequent waveform even if a portion of the waveform is segmented. Therefore, even if the frequency of the respiratory element changes midway, it is possible to track the original respiratory element. This method can also be applied to cardiovascular systems, such as the heartbeat, which can change suddenly. Next, the operation of each module according to this embodiment will be described.

[0136] [Respiratory function analysis] First, let's explain the respiratory function analysis. Figure 6A is a flowchart illustrating the overview of the respiratory function analysis according to this embodiment. Basic module 1 extracts DICOM images from database 15 (step S1). Here, at least multiple frame images contained within one respiratory cycle are acquired. Next, in each acquired frame image, the period of the respiratory element is identified using the density (density / intensity) in at least a certain region within the lung field (step S2). The identified respiratory cycle and the waveform identified from this respiratory cycle can be used in the following steps.

[0137] Identifying the periodicity of respiratory elements can also be done using diaphragmatic and thoracic movements. Alternatively, data obtained from other measurement methods, such as spirometry, can be used, including a range composed of a certain volume, "density," and "intensity" measured in areas with high X-ray transparency. It is also possible to pre-identify the frequencies associated with each organ (in this case, the lungs) and extract the "density" and "intensity" corresponding to those identified frequencies.

[0138] Next, in Figure 6A, the lung field is automatically detected (step S3). Since the lung contour changes continuously, if the maximum and minimum shapes can be detected, the shapes in between can be interpolated by calculation. Based on the period of the respiratory element identified in step S2, the lung contour in each frame image is identified by interpolating each frame image. Alternatively, the lung field may be detected by pattern matching as shown in Figures 2E to 2H. The detected lung field may also be subjected to noise reduction by cutoff. Next, the detected lung field is divided into multiple block regions (step S4). Then, the changes in each block region in each frame image are calculated (step S5). Here, the values ​​of change within each block region are averaged and expressed as a single data point.

[0139] Furthermore, noise reduction may be performed on the values ​​of change within each block region by cutoff. Next, Fourier analysis or synchronization rate analysis is performed on the "density" / "intensity" values ​​of each block region, as well as the amount of change thereof, based on the period of the respiratory element described above (Step S6).

[0140] Next, the results obtained from the Fourier analysis or the synchronization rate analysis are denoised (step S7). Here, cutoffs and artifacts can be removed as described above. The operations from steps S5 to S7 are performed one or more times, and it is determined whether the process is complete (step S8). Here, regarding the features displayed on the display, due to the mixing of composite waves and other waves, high-purity elements, such as respiratory elements, blood flow elements, and other elements, may not be displayed in a single spectral extraction. In such cases, the features displayed on the display may be used as pixel values, and all or part of the analysis leading to the display may be re-analyzed multiple times. This process makes it possible to obtain even higher-purity images regarding the synchronization and agreement of elements, such as respiratory elements and blood flow elements. This operation may be performed manually by the operator while visually observing the image on the display, or it may be performed automatically by extracting spectra from the output results and recalculating their distribution ratios. Furthermore, even after calculation, noise reduction, least squares interpolation, and correction using the surrounding "density" may be performed as needed.

[0141] If step S8 is not completed, the process proceeds to step S5. If it is completed, the results obtained from the Fourier analysis or synchronization rate analysis are displayed on the display as a pseudo-color image (step S9). Alternatively, a black and white image may be displayed. In this way, the accuracy of the data can be increased by repeating multiple cycles. This makes it possible to display the desired video. Alternatively, the desired video can be obtained by modifying the image displayed on the display.

[0142] In this embodiment, the desired frequency or frequency band is calculated, but this does not necessarily result in a good image when viewed as an actual image. Therefore, the following methods may be employed. (1) A method of presenting several frequency bands and allowing a human to make a selection. (2) A method of presenting several frequency bands and using AI technology to extract good images through pattern recognition. (3) Select based on the trend and shape of the histogram. That is, the value at the center of the histogram in the resulting signal tends to be high, and the histogram value fluctuates according to the movement, so you can also select based on the trend and shape of the histogram.

[0143] [Lung blood flow analysis] Next, we will explain pulmonary blood flow analysis. Figure 7 is a flowchart illustrating the overview of pulmonary blood flow analysis according to this embodiment. Basic module 1 extracts DICOM images from database 15 (step T1). Here, at least multiple frame images contained within one heartbeat cycle are acquired. Next, the vascular beat cycle is identified based on each acquired frame image (step T2). The identified vascular beat cycle and the waveform identified from this vascular beat cycle can be used in the following steps. As mentioned above, the vascular beat cycle is analyzed using, for example, measurement results from other modalities such as electrocardiograms and pulse meters, and changes in "density" / "intensity" of any part such as the heart, pulmonary hilum, and major blood vessels. Alternatively, the frequencies possessed by each organ (in this case, pulmonary blood flow) may be identified in advance, and the "density" / "intensity" corresponding to those identified frequencies may be extracted.

[0144] Next, in Figure 7, the period of the respiratory element is identified using the method described above (step T3), and the lung field is automatically detected using the period of the respiratory element (step T4). In the automatic detection of the lung contour, there may be variations from frame image to frame image, but the lung contour in each frame image is identified by interpolating each frame image based on the period of the respiratory element identified in step T3. Alternatively, the lung field may be detected by pattern matching as shown in Figures 2E to 2H. The detected lung field may also be subjected to noise reduction by cutoff. Next, the detected lung field is divided into multiple block regions (step T5). Then, the changes in each block region in each frame image are calculated (step T6). Here, the values ​​of change within each block region are averaged and expressed as a single data point. The values ​​of change within each block region may also be subjected to noise reduction by cutoff. Next, Fourier analysis or synchrony rate analysis is performed on the "density" / "intensity" values ​​of each block region, as well as the amount of change therein, based on the vascular rhythm described above (Step T7).

[0145] Next, the results obtained from the Fourier analysis or the synchronization rate analysis are denoised (step T8). Here, cutoffs and artifacts can be removed as described above. The operations from step T6 to step T8 are performed one or more times, and it is determined whether the process is complete (step T9). At this point, regarding the features displayed on the display, due to the presence of composite waves and other waves, high-purity elements, such as respiratory elements, blood flow elements, and other elements, may not be displayed in a single spectral extraction. In such cases, the features displayed on the display may be used as pixel values, and all or part of the analysis leading to the display may be re-analyzed multiple times. This process makes it possible to obtain even higher-purity images regarding the synchronization and agreement of elements, such as respiratory elements and blood flow elements. This operation may be performed manually by the operator while visually observing the image on the display, or it may be performed automatically by extracting spectra from the output results and recalculating their distribution ratios. Furthermore, even after calculation, noise reduction, least squares interpolation, and correction using the surrounding "density" may be performed as needed.

[0146] If step T9 is not completed, the process proceeds to step T6. If it is completed, the results obtained from the Fourier analysis or synchronization rate analysis are displayed on the display as a pseudo-color image (step T10). Alternatively, a black and white image may be displayed. This can improve the accuracy of the data. Furthermore, the desired video can be obtained by modifying the image displayed on the screen.

[0147] In this embodiment, the desired frequency or frequency band is calculated, but this does not necessarily result in a good image when viewed as an actual image. Therefore, the following methods may be employed. (1) A method of presenting several frequency bands and allowing a human to make a selection. (2) A method of presenting several frequency bands and using AI technology to extract good images through pattern recognition. (3) Select based on the trend and shape of the histogram. That is, the value at the center of the histogram in the resulting signal tends to be high, and the histogram value fluctuates according to the movement, so you can also select based on the trend and shape of the histogram.

[0148] [Other blood flow analyses] Next, other blood flow analyses will be described. One aspect of the present invention is applicable to the analysis of blood flow in the heart, aorta, pulmonary vessels, brachial artery, cervical vessels, etc., as shown in Figure 15. Furthermore, blood flow analysis can be similarly performed on abdominal vessels (not shown) and peripheral vessels. Figure 8 is a flowchart outlining other blood flow analyses according to this embodiment. Basic module 1 extracts DICOM images from database 15 (step R1). Here, at least multiple frame images contained within one heartbeat cycle are acquired. Next, the vascular beat cycle is identified based on each acquired frame image (step R2). The identified vascular beat cycle and the waveform identified from this vascular beat cycle can be used in the following steps. As described above, the vascular beat cycle is analyzed using, for example, measurement results from other modalities such as electrocardiograms and pulse meters, and changes in "density" / "intensity" at any site such as the heart, pulmonary hilum, and major vessels. Alternatively, the frequencies associated with each organ (for example, major blood vessels) can be identified in advance, and the corresponding "density" and "intensity" can be extracted.

[0149] Next, the analysis range is set (Step R3), and the set analysis range is divided into multiple block regions (Step R4). Then, the values ​​of change within each block region are averaged and represented as a single data point. Note that noise reduction may be performed on the values ​​of change within each block region using a cutoff. Next, Fourier analysis or synchronization rate analysis is performed on the "density" / "intensity" values ​​of each block region, as well as the amount of change, based on the vascular beat cycle described above (Step R5).

[0150] Next, denoising is performed on the results obtained from the Fourier analysis or the synchronization rate analysis (step R6). Here, cutoffs and artifact removal as described above can be performed. The operations from step R5 to step R6 are performed one or more times, and it is determined whether the process is complete (step R7). Here, regarding the features displayed on the display, due to the mixing of composite waves and other waves, high-purity elements, such as respiratory elements, blood flow elements, and other elements, may not be displayed in a single spectral extraction. In such cases, the features displayed on the display may be used as pixel values, and all or part of the analysis leading to the display may be re-analyzed multiple times. This process makes it possible to obtain even higher-purity images regarding the synchronization and agreement of elements, such as respiratory elements and blood flow elements. This operation may be performed manually by the operator while visually observing the image on the display, or it may be performed automatically by extracting spectra from the output results and recalculating their distribution ratios. Furthermore, even after calculation, noise reduction, least squares interpolation, and correction using the surrounding "density" may be performed as needed.

[0151] If step R7 is not completed, the process proceeds to step R5. If it is completed, the results obtained from the Fourier analysis or synchronization rate analysis are displayed on the screen as a pseudo-color image (step R8). Alternatively, a black and white image may be displayed. This can improve the accuracy of the data. Furthermore, the desired video can be obtained by modifying the image displayed on the screen.

[0152] In this embodiment, the desired frequency or frequency band is calculated, but this does not necessarily result in a good image when viewed as an actual image. Therefore, the following methods may be employed. (1) A method of presenting several frequency bands and allowing a human to make a selection. (2) A method of presenting several frequency bands and using AI technology to extract good images through pattern recognition. (3) Select based on the trend and shape of the histogram. That is, the value at the center of the histogram in the resulting signal tends to be high, and the histogram value fluctuates according to the movement, so you can also select based on the trend and shape of the histogram.

[0153] Furthermore, when analyzed in 3D, by measuring respiratory volume, cardiac output, and central blood flow with a separate device, it becomes possible to calculate respiratory volume, cardiac output, and central blood flow in each block region from the relative values ​​of the Fourier analysis results. In other words, in the case of respiratory function analysis, it becomes possible to estimate pulmonary ventilation volume from respiratory volume; in the case of pulmonary blood flow analysis, it becomes possible to estimate pulmonary blood flow volume from cardiac (pulmonary vascular) output; and in the case of other blood flow analyses, it becomes possible to estimate the estimated blood flow volume (percentage) in branched vessels depicted from the central blood flow volume (percentage).

[0154] Furthermore, as mentioned above, while calculating all data from the acquired database would allow for more accurate judgments, performing computer analysis can be time-consuming. Therefore, it is possible to extract only a specific number of frames (for example, a particular phase) and perform calculations on those frames. This shortens the analysis time and allows for the removal of irregular points, such as those observed at the beginning of respiration. Additionally, when displaying the analysis results, it is possible to display any range. For example, by displaying the range from the transition between "expiration / inspiration" to the transition between "inspiration / expiration," so-called "endless playback" becomes possible during repeated playback, making diagnosis easier for physicians.

[0155] As described above, this embodiment makes it possible to evaluate images of the human body using an X-ray video device. If digital data can be acquired, calculations can generally be performed well with existing facility equipment, resulting in low implementation costs. For example, in an X-ray video device using a flat panel detector, it becomes possible to easily examine the subject. Furthermore, for pulmonary blood flow, screening for pulmonary thromboembolism becomes possible. For example, in an X-ray video device using a flat panel detector, unnecessary examinations can be eliminated by running the diagnostic support program according to this embodiment before performing a CT scan. Also, because the examination is simple, it becomes possible to detect highly urgent diseases early and respond to them preferentially. Currently, there are several challenges with other modalities such as CT and MRI in imaging methods, but if these can be resolved, detailed diagnosis of each area will be possible.

[0156] Furthermore, it can be applied to screening various blood vessels, such as cervical blood flow narrowing, and also to evaluating and screening blood flow in large vessels. In addition, lung respiratory data can be used as a partial lung function test and can be used as a pulmonary function test. It can also be used to identify diseases such as COPD and emphysema. Moreover, it can be applied to understanding pre- and post-operative characteristics. Furthermore, by performing Fourier analysis on the respiratory element cycle and blood flow cycle, and removing the respiratory and blood flow waveforms from abdominal X-ray images, it becomes possible to observe remaining abnormalities in biological movement, such as intestinal ileus.

[0157] Note that if the initial image acquired is of a relatively high resolution, the calculation time may be longer due to the large number of pixels. In that case, you can reduce the image to a certain number of pixels before calculating. For example, you can reduce the calculation time by reducing a "4096 x 4096" image to "1024 x 1024" pixels before calculating.

[0158] [others] Furthermore, when acquiring X-ray images, predictive algorithms such as the AR method (Autoregressive Moving Average model) can be used. If at least one frequency of the respiratory element can be identified, it is also possible to control the X-ray imaging device to adjust the X-ray irradiation interval according to this frequency. For example, if the frequency of the respiratory element is low (long period), the number of X-ray imaging sessions can be reduced. This makes it possible to reduce the amount of radiation exposure to the human body. However, if the frequency of the respiratory or cardiovascular rhythm elements is high (short period), such as in cases of tachypnea or tachycardia, the irradiation frequency may be increased to create an optimal image.

[0159] Furthermore, regarding the storage format of DICOM data, it is preferable to save it uncompressed, as compression can sometimes degrade image quality. Alternatively, the calculation method may be changed depending on the data compression format. [Explanation of Symbols]

[0160] 1. Basic Module 3 Respiratory function analysis department 5 Pulmonary blood flow analysis department 7. Other blood flow analysis units 9. Fourier Analysis Section 10 Waveform analysis section 11 Visualization and Numericalization Section 13 Input Interfaces 15 Databases 17 Output Interface 19 displays

Claims

1. A diagnostic support program that analyzes images of the human body and outputs the analysis results, The process of acquiring multiple frame images, A process to calculate the difference or ratio between the pixel values ​​of at least one frame image and the pixel values ​​of at least one other frame image selected at an arbitrary interval, Using the calculated difference or ratio, the process involves identifying the lung field from the at least two frame images and calculating the pixel values ​​of the lung field region in the at least two frame images based on the density changes or the positional relationship of blood vessels in the lung field. A diagnostic support program characterized by comprising at least the process of visualizing the lung contour of any frame image within the interval between the at least two frame images.

2. The diagnostic support program according to claim 1, further comprising a process for calculating the pixel values ​​of a frame image from which artifacts have been removed.

3. A process for calculating the pixel values ​​within each frame image of the aforementioned plurality of frame images as relative or logarithmic values, A process for calculating the ratio between the relative value or logarithmic value of any frame image displayed as a relative value or logarithmic value and the relative value or logarithmic value of other frame images selected at arbitrary intervals and displayed as relative values ​​or logarithmic values, The process of visualizing the calculated ratio, A diagnostic support program according to claim 1 or 2, further comprising:

4. The diagnostic support program according to claim 1 or 2, characterized in that the process for calculating the difference or ratio divides the lung field into a plurality of block regions and calculates the average density of the block region in each frame image.

5. A diagnostic support program that analyzes images of the human body and outputs the analysis results, The process of acquiring multiple frame images, A process to identify the frequency of at least one respiratory element, including all or part of exhalation or inhalation, based on the pixels of a specific region of each frame image, A process to detect the lung contour using the identified frequency, From the detected lung contour, the process involves identifying at least one of the following: the heart, the pulmonary hilum, or a major blood vessel; and using the changes in density of at least one of the identified heart, pulmonary hilum, or major blood vessel, the process involves identifying the heartbeat or vascular beat. A diagnostic support program characterized by including a process of identifying at least one frequency or waveform of cardiovascular elements obtained from the identified heartbeat or vascular beat.

6. The process further includes outputting an image corresponding to the identified frequency or waveform. The diagnostic support program according to claim 5.

7. The diagnostic support program according to claim 5 or 6 is characterized by controlling the X-ray imaging apparatus to adjust the X-ray irradiation interval according to the frequency of at least one of the respiratory elements. Ram.

8. Based on the pixels of a specific region in each of the aforementioned frame images, multiple frequencies of respiratory elements, including all or part of exhalation or inhalation, are identified. The diagnostic support program according to claim 5 or 6, characterized in that it displays images corresponding to each of the multiple frequencies of the respiratory element on a display.

9. The diagnostic support program according to claim 5 or 6, further comprising the process of detecting the lung contour, identifying the diaphragm or thoracic cage, calculating the amount of change in the identified diaphragm or thoracic cage, and calculating the rate of change from the amount of change.

Images