Device, method, and program for calculating meandering degree

The meandering degree calculation device and method improve accuracy in evaluating blood vessel tortuosity by calculating angles between vectors in an image, addressing inaccuracies in existing methods and aligning closer to doctor's evaluations.

JP2024162468APending Publication Date: 2024-11-21CANADEVIA CO LTD +2
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
JP2023078001
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-10
Publication Date
2024-11-21

AI Technical Summary

Technical Problem

Existing methods for evaluating the degree of tortuosity of blood vessels, such as those described in Patent Documents 1 and 2, may deviate from actual doctor's findings, leading to inaccuracies in test results.

Method used

A meandering degree calculation device and method that calculates the meandering degree of an object by determining the angle between vectors formed by adjacent pixels in an image, using a first and second detection point, and calculating a statistical representation of these angles to enhance accuracy.

Benefits of technology

The device and method provide a more accurate calculation of the meandering degree, reducing discrepancies with actual doctor's findings and improving the reliability of test results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024162468000001_ABST
    Figure 2024162468000001_ABST
Patent Text Reader

Abstract

To achieve a meandering degree calculation device for calculating a meandering degree of an object with higher accuracy.SOLUTION: A meandering degree calculation device (1) includes a first calculation unit (134) and a second calculation unit (135). When different points on a line reflecting a meandering state of an image indicated by an image including an image of an object are defined as a first detection point P1 and a second detection point P2, the first calculation unit calculates an angle defined by a first vector connecting two adjacent pixels in a plurality of pixels constituting a line and a second vector connecting the first detection point P1 and the second detection point P2 about at least part of a set of a plurality of adjacent pixels from the first detection point P1 to the second detection point P2. The second calculation unit calculates, as a meandering degree between the first detection point P1 and the second detection point P2, a statistic indicating a feature of the plurality of angles calculated by the first calculation unit.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present disclosure relates to a meandering degree calculation device, a meandering degree calculation method, and a meandering degree calculation program. [Background technology]

[0002] Techniques for evaluating the tortuosity of blood vessels are known (Patent Documents 1 and 2). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 4-288124 [Patent Document 2] Japanese Patent Application Publication No. 5-123296 Summary of the Invention [Problem to be solved by the invention]

[0004] The techniques of Patent Documents 1 and 2 may cause discrepancies with the actual findings of a doctor, which may affect the accuracy of the examination. Further improvements in the quantitative evaluation method for the tortuosity of blood vessels are desired.

[0005] In view of the above-described problems, an aspect of the present disclosure has an object to provide a meandering degree calculation device or the like that calculates the meandering degree of an object with higher accuracy. [Means for solving the problem]

[0006] In order to solve the above problems, a meandering degree calculation device according to one embodiment of the present disclosure is a meandering degree calculation device that calculates the meandering degree of an object that extends in a meandering manner, and includes a first calculation unit that calculates, when different points on a line reflecting the meandering state of an image including an image of the object are set as first detection points and second detection points, an angle between a first vector connecting two adjacent pixels among multiple pixels constituting the line and a second vector connecting the first detection point and the second detection point for at least a portion of multiple pairs of adjacent pixels from the first detection point to the second detection point, and a second calculation unit that calculates a statistical quantity that indicates characteristics of the multiple angles calculated by the first calculation unit as the meandering degree between the first detection point and the second detection point.

[0007] In order to solve the above problems, a meandering degree calculation method according to one embodiment of the present disclosure is a method for calculating the meandering degree of an object extending in a meandering manner, the method including: a first step of calculating, when different points on a line reflecting the meandering state of an image including an image of the object are defined as a first detection point and a second detection point, an angle between a first vector connecting two adjacent pixels among a plurality of pixels constituting the line and a second vector connecting the first detection point and the second detection point, for at least a portion of a plurality of pairs of adjacent pixels from the first detection point to the second detection point; and a second step of calculating, as the meandering degree between the first detection point and the previous second detection point, a statistical quantity indicating characteristics of the plurality of angles calculated in the first step. Effect of the Invention

[0008] According to one aspect of the present disclosure, it is possible to realize a meandering degree calculation device or the like that calculates the meandering degree of an object with higher accuracy. [Brief description of the drawings]

[0009] [Figure 1] 1 is a functional block diagram illustrating an embodiment of a meandering degree calculation device according to the present disclosure. [Diagram 2]4 is a flowchart showing a process flow of the meandering degree calculation device according to the present disclosure. [Diagram 3] FIG. 4 is a diagram for explaining the operation of a first calculation unit. [Figure 4] FIG. 1 is a diagram for explaining an example of determining the degree of meandering according to a conventional method. [Diagram 5] The three photographs used in the comparative test are shown below. [Figure 6] FIG. 13 is a diagram showing the results of a comparison test. [Figure 7] FIG. 13 is a diagram showing the relationship between a doctor's findings and evaluation scores for seven blood vessel images. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, a meandering degree calculation device 1, a meandering degree calculation method, and a meandering degree calculation program according to the present disclosure will be described with reference to the drawings.

[0011] First, an overview of the meandering degree calculation device 1 according to the present disclosure will be given. The meandering degree calculation device 1 includes a first calculation unit 134 that calculates, when different points on a line reflecting the meandering state of an image shown by an image including an image of an object are defined as a first detection point P1 and a second detection point P2, an angle between a first vector connecting two adjacent pixels among multiple pixels constituting the line and a second vector connecting the first detection point P1 and the second detection point P2 for at least a portion of multiple pairs of adjacent pixels from the first detection point P1 to the second detection point P2, and a second calculation unit 135 that calculates a statistical quantity indicating characteristics of the multiple angles calculated by the first calculation unit as the meandering degree between the first detection point P1 and the second detection point P2.

[0012] In the following embodiment, the object is assumed to be a blood vessel contained in a fundus image. "Blood vessel" includes not only human blood vessels but also blood vessels of animals other than humans. The object may be a tangible object that extends in a meandering manner (such as a hose) or an intangible object such as an image showing the trajectory of a car or the like. "Meandering" refers to a state in which the body moves in a twisting manner like a snake, but in a broader sense, may also include a state in which the body moves in a straight line. "Statistics" refers to the first vector (V i The angle between the first vector (P) and the second vector (P) is called the angle θ i When (described in detail later), it may be the variance of cos θ, the degree of change of cos θ along the path between two points, or the angular variance of θ, but in the following embodiment, it is the angular variance of θ.

[0013] First, a background in which blood vessels contained in a fundus image are the subject will be described.

[0014] For early detection of vascular disorders in the brain, medical equipment that visualizes the inside of the human body, such as CT (Computed Tomography) and MRI (Magnetic Resonance Imaging), is used. The brain is the same organ as the eye in terms of development. Therefore, the state of blood vessels in the brain can be inferred from the state of retinal blood vessels in the fundus. For risk diagnosis of vascular disorders in the brain, fundus examination, which does not require the aforementioned medical equipment, has been used. In recent years, fundus examination has been incorporated into internal medicine examination items due to the revision of the third stage specific health checkup. Fundus examination can be used not only for prevention or early detection of eye diseases, but also for estimating the progress of arteriosclerosis in blood vessels in the brain. Arteriosclerosis does not progress at a constant rate, but at an accelerated rate. Therefore, early treatment of arteriosclerosis is performed by observing the degree of arteriosclerosis over time and finding patients whose arteriosclerosis is accelerating.

[0015] On the other hand, fundus examinations depend heavily on the experience and skill of the examiner. Therefore, it is difficult for non-specialist physicians to perform accurate fundus examinations. Therefore, there is a demand for automation of fundus examinations.

[0016] The tortuosity calculation device 1 according to the present disclosure includes the configuration described below, and can calculate the tortuosity of blood vessels included in a fundus image with high accuracy. The tortuosity of blood vessels is useful information for understanding the condition of arteriosclerosis, hypertension, and the like. Furthermore, the tortuosity calculation device 1 can calculate an index value indicating the presence or absence of an abnormality in blood vessels. The tortuosity calculation device 1 will be described below.

[0017] (Meandering degree calculation device) FIG. 1 is a functional block diagram showing an embodiment of a meandering degree calculation device 1. As shown in FIG.

[0018] The meandering degree calculation device 1 is connected to the imaging device 2 via a network (communication line) that uses a wired communication method or a wireless communication method. The network is, for example, a communication network such as the Internet, a mobile communication network, a Local Area Network (LAN), a Wide Area Network (WAN), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0019] The imaging device 2 captures an image. The captured image includes, for example, a fundus image captured by a fundus camera, an X-ray image captured by an X-ray imaging device, a CT image captured by a CT device, and an image captured by a normal optical camera. In the present disclosure, at least blood vessels are captured as subjects in the captured image.

[0020] The meandering degree calculation device 1 is realized, for example, by a personal computer. A meandering degree calculation program runs in the meandering degree calculation device 1, and the meandering degree calculation program cooperates with the hardware resources of the meandering degree calculation device 1 to realize a meandering degree calculation method.

[0021] The meandering degree calculation program causes a computer (not shown) to function in the same manner as the meandering degree calculation device 1 by executing the meandering degree calculation program, thereby causing the computer to execute the meandering degree calculation method.

[0022] The meandering degree calculation device 1 includes a communication unit 11, a storage unit 12, and a control unit 13.

[0023] The communication unit 11 acquires the captured image from the imaging device 2 via a network. The communication unit 11 is configured with, for example, a communication module and an antenna. The captured image is stored in the storage unit 12.

[0024] The storage unit 12 stores information necessary for the meandering calculation device 1 to execute the meandering calculation method. The storage unit 12 is configured, for example, with a recording device such as a hard disk drive (HDD) or a solid state drive (SSD) provided in the meandering calculation device 1, a volatile memory such as a random access memory (RAM), and / or a portable storage medium such as a flash memory.

[0025] The control unit 13 controls the overall operation of the meandering calculation device 1 and executes the meandering calculation method. The control unit 13 is configured, for example, with a CPU (Central Processing Unit) provided in the meandering calculation device 1, a volatile memory such as a RAM that functions as a working area for the CPU, and a non-volatile memory such as a ROM (Read Only Memory) that stores various information such as a meandering calculation program.

[0026] The control unit 13 includes an extraction image generating unit 131 , a thin line image generating unit 132 , a detection unit 133 , a first calculation unit 134 , a second calculation unit 135 , and a third calculation unit 136 .

[0027] The extracted image generation unit 131, the thinned image generation unit 132, the detection unit 133, the first calculation unit 134, the second calculation unit 135, and / or the third calculation unit 136 may be configured to include hardware such as a GPU (Graphics Processing Unit), a DSP (Digital Signal Processor), or an FPGA (Field Programmable Gate Array) in addition to or instead of a CPU.

[0028] The extracted image generation unit 131, the thinned image generation unit 132, the detection unit 133, the first calculation unit 134, the second calculation unit 135, and / or the third calculation unit 136 may be configured using common hardware (such as a CPU) or may be configured using individual hardware.

[0029] Hereinafter, the configuration and operation of each part of the control unit 13 will be described with reference to Fig. 1 and Fig. 2. Fig. 2 is a flowchart showing the flow of processing of the meandering degree calculation device 1 according to the present disclosure.

[0030] The extracted image generating unit 131 generates an extracted image based on a captured image (S10 in FIG. 2). The extracted image is an image obtained by extracting a blood vessel region included in the captured image.

[0031] Specifically, the extracted image generation unit 131 reads out the captured image from the memory unit 12, and performs known image processing on the captured image (for example, processing using a Convolutional Neural Network (CNN) using learning, or processing using a Black-Top-Hat (BTH) transform, which is a type of morphological operation) to extract blood vessel areas included in the captured image and generate an extracted image. For example, the extracted image generation unit 131 binarizes blood vessel areas included in the captured image into white and areas other than the blood vessel areas (background areas) into black, thereby generating an extracted image in which the captured image is binarized. The extracted image is stored in the memory unit 12.

[0032] The thin line image generating unit 132 generates a thin line image based on the extracted image (S20 in FIG. 2). The thin line image is an image in which the blood vessel region included in the extracted image is thinned to a width of one pixel. In other words, the thin line image is an image in which the blood vessels included in the captured image are thinned.

[0033] The thinned image generating unit 132 performs a known thinning process on the extracted image to thin the blood vessel area included in the extracted image to a width of one pixel, thereby generating a thinned image. The thinned image is stored in the storage unit 12.

[0034] The known thinning process is an image process for thinning white areas of a binarized image, and is, for example, the process described in the following document: “Zhang TY, and Ching Y. Suen, “A fast parallel algorithm for thinning digital patterns”, Communications of the ACM, (US), 1984, Vol. 27, No. 3, pp. 236-239”.

[0035] The detection unit 133 detects a plurality of detection points corresponding to any of the intersections, branching points, and end points of the thinned blood vessels based on the thinned image (S30 in FIG. 2).

[0036] As an example, the detection unit 133 applies each pattern of the filter to each pixel of the thinned image to identify pixels that are detection points. That is, as an example, the detection unit 133 identifies detection points by performing a filter process using a filter on the thinned image.

[0037] The filter is a filter that is applied to the thinned image so that the detection unit 133 can identify crossing points, branching points, end points, and passing points of blood vessels from each pixel of the thinned image. The filter includes a plurality of patterns corresponding to crossing points, branching points, end points, and passing points of blood vessels, respectively. The filter is stored in the storage unit 12 in advance.

[0038] The detection unit 133 searches for thinned blood vessels connected to each detection point based on the thinned image and the coordinate information of the detection points associated with the thinned image. The detection unit 133 then identifies a pair of two detection points connected by a thinned blood vessel among the detected multiple detection points. The detection point pair information is stored in the storage unit 12.

[0039] The above describes the extraction image generation unit 131, the thinned image generation unit 132, and the detection unit 133. The processing by the extraction image generation unit 131, the thinned image generation unit 132, and the detection unit 133 may be processing using publicly known technology, and one example is the processing described in a patent application by the present applicant (Patent Application No. 2021-26551).

[0040] The meandering degree calculation device 1 may not include the extracted image generation unit 131, the thinned image generation unit 132, and the detection unit 133. The meandering degree calculation device 1 may cause an external device (not shown) to execute the processes to be executed by the extracted image generation unit 131, the thinned image generation unit 132, and the detection unit 133, and may acquire the results of the execution via the communication unit 11.

[0041] Furthermore, the processes executed by the extraction image generating unit 131, the thinned image generating unit 132, and / or the detection unit 133 may be performed using other known image processing techniques, and are not limited to a specific image processing technique.

[0042] (First calculation section) Next, the first calculation unit 134 will be described.

[0043] First, a pair of detection points (hereinafter referred to as "first detection point P1" and "second detection point P2") are extracted from the multiple detection points detected by the detection unit 133. The first detection point P1 and the second detection point P2 are different points on a line that reflects the meandering state of the blood vessel shown in an image (thinned image) including an image of the blood vessel. The first detection point P1 and the second detection point P2 may be extracted manually by a doctor or the like, or may be extracted randomly.

[0044] Here, a vector connecting two adjacent pixels among a plurality of pixels constituting a line reflecting the tortuous state of the blood vessel shown in the thinned image is referred to as a first vector, and a vector connecting the first detection point P1 and the second detection point P2 is referred to as a second vector. The first calculation unit 134 calculates the angle between the first vector and the second vector for at least a portion of the pairs of adjacent pixels from the first detection point P1 to the second detection point P2 (S40 in FIG. 2).

[0045] Hereinafter, details of the operation of the first calculation unit 134 will be described with reference to Fig. 3. Fig. 3 is a diagram for explaining the operation of the first calculation unit 134.

[0046] FIG. 3 shows an image in which blood vessel regions included in an extracted image have been thinned to a width of one pixel. In FIG. 3, pixels shown in gray are pixels that represent thinned blood vessels. Pixels that represent areas other than thinned blood vessels are shown in white. In FIG. 3, P1 represents the first detection point P1. P2 represents the second detection point P2. P represents a second vector connecting the first detection point P1 and the second detection point P2.

[0047] Reference numeral 301 in Fig. 3 is a diagram explaining a first vector (V0) connecting the 0th pixel from the first detection point P1 (i.e., the first detection point P1) and the first pixel from the first detection point P1. Reference numeral 302 in Fig. 3 is a diagram explaining a first vector (V1) connecting the first pixel from the first detection point P1 and the second pixel from the first detection point P1. Reference numeral 303 in Fig. 3 is a diagram explaining a first vector (V2) connecting the second pixel from the first detection point P1 and the third pixel from the first detection point P1.

[0048] 3, V0 denotes a first vector connecting the 0th pixel from the first detection point P1 (i.e., the first detection point P1) and the first pixel from the first detection point P1. The first calculation unit 134 calculates the angle (θ0) between the first vector (V0) and the second vector (P).

[0049] 3, V1 denotes a first vector connecting the first pixel from the first detection point P1 to the second pixel from the first detection point P1. The first calculation unit 134 calculates the angle (θ1) between the first vector (V1) and the second vector (P).

[0050] 3, V2 denotes a first vector connecting the second pixel from the first detection point P1 and the third pixel from the first detection point P1. The first calculation unit 134 calculates the angle (θ2) between the first vector (V2) and the second vector (P).

[0051] In this manner, the first calculation unit 134 calculates the angle between the first vector and the second vector for each of a plurality of pairs of adjacent pixels from the first detection point P1 to the second detection point P2. The first calculation unit 134 may calculate the angle between the first vector and the second vector for at least some of the plurality of pairs of adjacent pixels from the first detection point P1 to the second detection point P2.

[0052] The first calculation unit 134 calculates the first vector (V i ) and the second vector (P) i ) is calculated based on the following formula (1).

number

number

[0053] Example 1 Next, the results of a comparison between the tortuosity calculation device 1 according to the present disclosure and the tortuosity determination according to the conventional method will be described with reference to FIGS. 4 to 6 using three blood vessel images.

[0054] (Determination of meandering degree according to conventional method) First, the determination of tortuosity according to the conventional method will be described. FIG. 4 is a diagram for explaining an example of the determination of tortuosity according to the conventional method. FIG. 4 shows an image in which the blood vessel region included in the extracted image has been thinned to a width of one pixel. In FIG. 4, pixels shown in gray are pixels that represent the thinned blood vessels. Pixels that represent other than the thinned blood vessels are shown in white. Reference numeral 401 in FIG. 4 indicates a thinned image in the case where the blood vessel tortuosity is deemed to be low (hereinafter referred to as "Case 1"). Reference numeral 402 in FIG. 4 indicates a thinned image in the case where the blood vessel tortuosity is deemed to be high (hereinafter referred to as "Case 2").

[0055] The conventional method determines the tortuosity of a blood vessel based on the ratio (e1 / e0) of the minimum number of pixels (e0) between the first detection point P1 and the second detection point P2 and the number of blood vessel pixels (e1). Comparing Case 1 and Case 2, the minimum number of pixels (e0) between the first detection point P1 and the second detection point P2 is the same, but the number of blood vessel pixels (e1) is greater in Case 2 than in Case 1. Therefore, the ratio (e1 / e0) is greater in Case 2, and it is determined that the tortuosity is greater in Case 2 than in Case 1.

[0056] In this way, the conventional method determines the degree of tortuosity according to the ratio (e1 / e0). Below, the results of a comparison between the conventional method and the method according to the present disclosure will be described.

[0057] (Comparison result 1) FIG. 5 shows three photographs used in the comparative test. In FIG. 5, the blood vessels used in the evaluation of tortuosity are shown in white. Reference numeral 501 is a photograph showing a blood vessel with low tortuosity (hereinafter referred to as "blood vessel 1"). Reference numeral 502 is a photograph showing a blood vessel with medium tortuosity (hereinafter referred to as "blood vessel 2"). Reference numeral 503 is a photograph showing a blood vessel with high tortuosity (hereinafter referred to as "blood vessel 3").

[0058] 6 is a diagram showing the results of a comparative test. The upper column shows the ratio (e1 / e0) calculated according to a conventional method. The lower column shows the angle dispersion calculated by the meandering degree calculation device 1 according to the present disclosure.

[0059] Looking at the upper column of Figure 6, the values ​​for Blood Vessel 1, Blood Vessel 2, and Blood Vessel 3 are "1.082," "1.088," and "1.535," respectively, and the values ​​for Blood Vessel 1 and Blood Vessel 2 are almost the same. However, comparing Blood Vessel 1 and Blood Vessel 2 in Figure 6, Blood Vessel 2 appears to have a greater tortuosity than Blood Vessel 1. For this reason, determining tortuosity according to the conventional method may cause a discrepancy between the calculated ratio (e1 / e0) and the doctor's actual findings, raising concerns about the impact on examination accuracy.

[0060] On the other hand, when checking the lower column of Fig. 6, the values ​​for Blood Vessel 1, Blood Vessel 2, and Blood Vessel 3 are "0.085", "0.146", and "0.236", respectively, and the values ​​increase as one progresses from Blood Vessel 1 to Blood Vessel 2 to Blood Vessel 3. This result shows a similar tendency to the findings obtained from the three photographs in Fig. 5. From the above results, it can be seen that the tortuosity calculation device 1 according to the present disclosure calculates the tortuosity of an object with higher accuracy than the conventional method, and therefore is less likely to cause a discrepancy with the actual findings of a doctor.

[0061] (Comparison result 2) Next, the comparison result 2 will be described with reference to FIG. 7. FIG. 7 is a diagram showing the relationship between the doctor's findings and the evaluation value for seven blood vessel images (not shown). The diagram indicated by reference numeral 701 in FIG. 7 is a graph showing the numbers of blood vessel images arranged by a doctor in ascending order of tortuosity according to the conventional method, on the horizontal axis, and the evaluation value (ratio (e1 / e0)) of each blood vessel image on the vertical axis. The diagram indicated by reference numeral 702 in FIG. 7 is a graph showing the numbers of blood vessel images arranged by a doctor in ascending order of tortuosity according to the conventional method, on the horizontal axis, and the evaluation value (angle variance) of each blood vessel image on the vertical axis, on the tortuosity calculated by the tortuosity calculation device 1 according to the present disclosure. In the two graphs in FIG. 7, the higher the value, the more consistent the evaluation by the doctor is.

[0062] First, looking at the graph indicated by reference numeral 701, the ratio (e1 / e0) of the blood vessel pointed out by the doctor to be the fourth from the bottom is about 1.15. This value is higher than the ratios of the blood vessels pointed out by the doctor to be the fifth and sixth from the bottom. In other words, for the blood vessels pointed out by the doctor to be the fourth to sixth from the bottom, there is a discrepancy between the ratio (e1 / e0) and the doctor's findings.

[0063] Next, looking at the graph indicated by reference numeral 702, the angle variance of the blood vessel that the doctor pointed out to be the fourth from the bottom is about 0.125. This value is higher than the angle variance of the blood vessel that the doctor pointed out to be the fifth from the bottom. In other words, for the blood vessels that the doctor pointed out to be the fourth and fifth from the bottom, there is a discrepancy between the angle variance and the doctor's findings.

[0064] From the above results, it is evident that the meandering degree calculation device 1 according to the present disclosure indicates an evaluation that is closer to the doctor's evaluation than the meandering degree judgment according to the conventional method, with less deviation from the doctor's findings.

[0065] From the above results, it can be seen that the tortuosity calculation device 1 according to the present disclosure calculates the tortuosity of an object with higher accuracy than the conventional method, and therefore is less likely to cause discrepancy with the actual findings of a doctor.

[0066] (Third calculation section) Next, the third calculation unit 136 will be described. As shown in FIG. 1, the tortuosity calculation device 1 according to the present disclosure may include the third calculation unit 136. The third calculation unit 136 calculates an index value indicating the presence or absence of a blood vessel abnormality based on the angle variance calculated from the above-mentioned formula (2). For example, the third calculation unit 136 calculates the ratio (V / n) of the angle variance (V) calculated by the second calculation unit 135 to a predetermined threshold value n as an index value indicating the presence or absence of a blood vessel abnormality. As an example, when V / n is larger than 1, it indicates that there is an abnormality in the blood vessel. When V / n is 1 or less, it indicates that there is no abnormality in the blood vessel. In this way, a doctor can diagnose whether there is an abnormality in the patient's blood vessel based on the quantitative index V / n.

[0067] The predetermined threshold value n may be recorded in the storage unit 12. The predetermined threshold value n may be changed as appropriate by a doctor or the like.

[0068] Furthermore, the third calculation unit 136 may calculate an index value based on a change over time in the angle variance calculated by the second calculation unit 135 using a plurality of segmented images obtained from a plurality of fundus images over time related to a specific individual. For example, the third calculation unit 136 calculates a ratio (V / n) of the angle variance (V) calculated by the second calculation unit 135 to a predetermined threshold n over time as an index value indicating the presence or absence of a blood vessel abnormality. When V / n is greater than 1, it indicates that there is an abnormality in the blood vessel. When V / n is equal to or less than 1, it indicates that there is no abnormality in the blood vessel. Alternatively, the third calculation unit 136 may calculate a rate of change over time of V / n. The third calculation unit 136 judges whether or not the abnormality in the blood vessel is progressing based on whether or not the rate of change over time of V / n exceeds a predetermined threshold. These are just examples, but in this way, a doctor can diagnose whether or not there is an abnormality in the blood vessel of a specific individual (patient), whether or not the abnormality is progressing, etc., based on the calculation result by the third calculation unit 136.

[0069] The third calculation unit 136 may calculate the index value based on the angle variance calculated by the second calculation unit 135 using a plurality of segmented images obtained from fundus images of a plurality of people related to a specific group to which the plurality of people belong. For example, the third calculation unit 136 calculates an average value of the angle variance using a plurality of segmented images obtained from fundus images of a plurality of people. When the angle variance of a specific individual is larger than the average value, a doctor can alert the specific individual to the possibility of an abnormality.

[0070] The third calculation unit 136 may calculate the index value based on the angle variance calculated by the second calculation unit 135 using a plurality of segmented images obtained from fundus images of a plurality of people related to a specific group to which the plurality of people belong. For example, the third calculation unit 136 calculates an average value of the angle variance over time using a plurality of segmented images obtained from fundus images of a plurality of people over time. When the rate of change of the angle variance of a specific individual over time is larger than the rate of change of the average value, a doctor can alert the specific individual to the possibility of an abnormality.

[0071] In this way, by including the third calculation unit 136, the meandering calculation device 1 according to the present disclosure can provide a doctor or the like with information for determining the presence or absence of an abnormality in an object based on a more quantitative index.

[0072] 〔summary〕 A meandering degree calculation device according to aspect 1 of the present disclosure is a meandering degree calculation device that calculates the meandering degree of an object that extends in a meandering manner, and includes a first calculation unit that calculates, when different points on a line reflecting the meandering state of an image including an image of the object are set as first detection points and second detection points, an angle between a first vector connecting two adjacent pixels among multiple pixels constituting the line and a second vector connecting the first detection point and the second detection point for at least a portion of multiple pairs of adjacent pixels from the first detection point to the second detection point, and a second calculation unit that calculates a statistical quantity indicating characteristics of the multiple angles calculated by the first calculation unit as the meandering degree between the first detection point and the second detection point.

[0073] According to the above configuration, the meandering degree calculation device according to the first aspect of the present disclosure can calculate the meandering degree of the object with higher accuracy.

[0074] In the meandering degree calculation device according to the second aspect of the present disclosure, in the first aspect, the first calculation unit calculates the angle based on the following formula (1):

number

number

[0075] The meandering degree calculation device according to a third aspect of the present disclosure is the device of the first or second aspect, further comprising a third calculation unit that calculates an index value indicating the presence or absence of an abnormality in the object based on the statistics.

[0076] According to the above configuration, it is possible to determine the presence or absence of an abnormality in the object based on a more quantitative index that is based on the index value calculated by the third calculation unit.

[0077] A tortuosity calculation device according to a fourth aspect of the present disclosure is any one of the third aspects, wherein the object is a blood vessel included in a fundus image.

[0078] Fundus examinations can be used not only to prevent or detect eye diseases early, but also to estimate the degree of progression of arteriosclerosis in blood vessels in the brain. Therefore, by calculating the tortuosity of blood vessels contained in a fundus image, it is possible to grasp the condition of arteriosclerosis, hypertension, etc.

[0079] A meandering degree calculation device according to aspect 5 of the present disclosure is configured in such a manner that, in the above-mentioned aspect 4, the third calculation unit calculates the index value based on the change over time in the statistical quantity calculated by the second calculation unit using a plurality of images obtained from a plurality of fundus images over time relating to a specific individual.

[0080] According to the above configuration, for example, patients in whom the progression of arteriosclerosis is accelerating can be easily found based on the index value.

[0081] A meandering degree calculation method according to aspect 6 of the present disclosure is a method for calculating the meandering degree of an object extending in a meandering manner, comprising: a first step of calculating, when different points on a line reflecting the meandering state of an image including an image of the object are defined as a first detection point and a second detection point, an angle between a first vector connecting two adjacent pixels among multiple pixels constituting the line and a second vector connecting the first detection point and the second detection point for at least a portion of multiple pairs of adjacent pixels from the first detection point to the second detection point; and a second step of calculating, as the meandering degree between the first detection point and the previous second detection point, a statistical quantity indicating characteristics of the multiple angles calculated in the first step.

[0082] According to the above configuration, the meandering degree calculation method according to the sixth aspect of the present disclosure can calculate the meandering degree of the object with higher accuracy.

[0083] A meandering calculation program according to aspect 7 of the present disclosure is a meandering calculation program for causing a computer to function as a meandering calculation device according to any one of aspects 1 to 5, and causes a computer to function as the first calculation unit and the second calculation unit.

[0084] According to the above configuration, the meandering degree calculation program according to the seventh aspect of the present disclosure can calculate the meandering degree of an object with higher accuracy.

[0085] The present disclosure is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present disclosure. [Explanation of symbols]

[0086] 1. Meandering degree calculation device 2. Imaging device 11 Communications Department 12 Storage section 13 Control section 131 Extraction image generation unit 132 Thinning image generation unit 133 Detection Unit 134 First Calculation Section 135 Second Calculation Section 136 Third Calculation Section P1 First detection point P2 Second detection point

Claims

1. A meandering degree calculation device for calculating a meandering degree of an object extending in a meandering manner, comprising: When different points on a line reflecting a meandering state of the image including the image of the object are defined as a first detection point and a second detection point, a first calculation unit that calculates an angle between a first vector connecting two adjacent pixels among a plurality of pixels constituting the line and a second vector connecting the first detection point and the second detection point for at least a portion of a plurality of pairs of adjacent pixels from the first detection point to the second detection point; a second calculation unit that calculates a statistic indicating a characteristic of the plurality of angles calculated by the first calculation unit as a degree of meandering between the first detection point and the second detection point; A meandering degree calculation device comprising:

2. The first calculation unit calculates the angle based on the following formula (1), [0010] P: Second vector V i : the first vector connecting the i-th pixel and the (i+1)-th pixel from the first detection point The meandering degree calculation device according to claim 1 , wherein the second calculation unit calculates the angle variance as the statistical amount based on the following formula (2): [0025] V: Angular dispersion R: average resultant vector length n: the number of the first vectors

3. The meandering degree calculation device according to claim 1 , further comprising a third calculation unit that calculates an index value indicating the presence or absence of an abnormality in the object based on the statistics.

4. The tortuosity calculation device according to claim 3 , wherein the object is a blood vessel included in a fundus image.

5. 5. The meandering degree calculation device according to claim 4, wherein the third calculation unit calculates the index value based on a change over time in the statistical amount calculated by the second calculation unit using a plurality of images obtained from a plurality of fundus images over time relating to a specific individual.

6. A method for calculating a meandering degree of an object extending in a meandering manner, comprising the steps of: When different points on a line reflecting a meandering state of the image including the image of the object are defined as a first detection point and a second detection point, a first step of calculating an angle between a first vector connecting two adjacent pixels among a plurality of pixels constituting the line and a second vector connecting the first detection point and the second detection point for at least a portion of a plurality of pairs of adjacent pixels from the first detection point to the second detection point; a second step of calculating a statistic indicating characteristics of the plurality of angles calculated in the first step as a degree of meandering between the first detection point and the second detection point; A method for calculating the meandering degree, comprising:

7. A meandering degree calculation program for causing a computer to function as the meandering degree calculation device according to claim 1, the program causing a computer to function as the first calculation unit and the second calculation unit.

Citation Information

Patent Citations

  • Method and apparatus for analyzing image of eyeground

    JP1992288124A

  • Image processor

    JP1993123296A