Method and apparatus for analyzing blood vessel based on machine learning model

Patent metadata

Field	Value
Publication number	US-12591966-B2
Application number	US-202318209389-A
Country	US
Kind code	B2
Filing date	Jun 13, 2023
Priority date	Oct 24, 2022
Publication date	Mar 31, 2026
Grant date	Mar 31, 2026

Abstract

An apparatus for analyzing a blood vessel includes a processor configured to detect a blood vessel area corresponding to the blood vessel and a stenosis area corresponding to a stenosed portion of the blood vessel, based on a machine learning model, from a blood vessel image, determine a partial area of the blood vessel area to be a comparison area for the stenosis area, based on a length of the stenosis area, and calculate a stenosis score indicating a degree of stenosis of a blood vessel of the stenosis area compared to a blood vessel of the comparison area.

First claim

What is claimed is:

1 . A method performed by an apparatus for analyzing a blood vessel, the method comprising:
receiving a blood vessel image from an input unit; preprocessing, with a processor, the blood vessel image to determine a target blood vessel of blood vessels in the blood vessel image, wherein the target blood vessel in the blood vessel image is emphasized by blurring surrounding structure in the blood vessel image; detecting, with the processor a blood vessel area corresponding to the target blood vessel and a stenosis area corresponding to a stenosed portion of the target blood vessel, based on a machine learning model, from the blood vessel image; determining, with the processor and the machine learning model, a partial area of the blood vessel area to be a comparison area for the stenosis area, based on a length of the stenosis area, by determining a first area and a second area each on respective sides of the stenosis area in a longitudinal direction of the target blood vessel to be the comparison area, wherein a median value of a blood vessel diameter of the second area is less than a median value of a blood vessel diameter of the first area; and calculating, with the processor, a stenosis score indicating a degree of stenosis of a blood vessel of the stenosis area compared to a blood vessel of the comparison area, based on a distribution of blood vessel diameters in the longitudinal direction of the target blood vessel in the blood vessel area.

2 . The method of claim 1 , wherein the detecting the blood vessel area and the stenosis area comprises:
detecting an area comprising a blood vessel diameter decreasing in the longitudinal direction of the target blood vessel as the comparison area in an area different from the stenosis area of the blood vessel area.

3 . The method of claim 1 , wherein the determining the comparison area comprises:
determining a partial centerline based on a partial reference line corresponding to the stenosis area of a reference line of the target blood vessel; and determining an area corresponding to the determined partial centerline to be the comparison area.

4 . The method of claim 1 , wherein the determining the comparison area further comprises:
determining the comparison area corresponding to a normal blood vessel adjacent to the stenosed portion of the stenosis area of the blood vessel area.

5 . The method of claim 1 , wherein determining the comparison area further comprises:
determining the first area corresponding to a more proximal blood vessel than the target blood vessel of the stenosis area and the second area corresponding to a more distal blood vessel than the target blood vessel of the stenosis area to be the comparison area.

6 . The method of claim 1 , wherein the calculating the stenosis score comprises:
extracting blood vessel diameters corresponding to the stenosis area and the comparison area from the distribution of the blood vessel diameters; and calculating the stenosis score by using the extracted blood vessel diameters.

7 . The method of claim 1 , wherein the calculating the stenosis score further comprises:
determining a reference vessel diameter estimated as a diameter of a normal blood vessel in the stenosis area, based on the distribution of the blood vessel diameters in the comparison area; determining a stenosed blood vessel diameter calculated in the stenosis area, based on the distribution of the blood vessel diameters in the stenosis area; and calculating the stenosis score based on a comparison result of the reference vessel diameter and the stenosed blood vessel diameter.

8 . The method of claim 1 , wherein the calculating the stenosis score further comprises:
extracting a first blood vessel diameter from a partial distribution for the first area of the comparison area of the distribution of the blood vessel diameters; extracting a second blood vessel diameter from a partial distribution for the second area of the comparison area of the distribution of the blood vessel diameters; calculating a reference vessel diameter corresponding to a normal blood vessel, based on the first blood vessel diameter and the second blood vessel diameter; extracting a minimum value of a partial distribution for the stenosis area of the distribution of the blood vessel diameters as a stenosed blood vessel diameter corresponding to the stenosis area; and calculating the stenosis score based on a comparison result of the calculated reference vessel diameter and the extracted stenosed blood vessel diameter.

9 . The method of claim 1 , further comprising:
detecting a calcification area of the blood vessel area based on a pixel value of the blood vessel image; and calculating a calcification score indicating a degree of calcification of the target blood vessel, based on a ratio of the calcification area to the blood vessel area.

10 . The method of claim 9 , further comprising outputting a graphic representation indicating the calcification area to the blood vessel image.

11 . The method of claim 1 , further comprising detecting a plaque area corresponding to a vulnerable plaque, together with the blood vessel area, by applying the machine learning model to the blood vessel image.

12 . An apparatus for analyzing a blood vessel, the apparatus comprising:
an input unit; and a processor configured to preprocess a blood vessel image received by the input unit to determine a target blood vessel of blood vessels in the blood vessel image, wherein the target blood vessel in the blood vessel image is emphasized by blurring surrounding structure in the blood vessel image, wherein the processor is configured to detect a blood vessel area corresponding to the target blood vessel and a stenosis area corresponding to a stenosed portion of the target blood vessel, based on a machine learning model, from the blood vessel image, determine with the machine learning model a partial area of the blood vessel area to be a comparison area for the stenosis area, based on a length of the stenosis area by determining a first area and a second area each on respective sides of the stenosis area in a longitudinal direction of the target blood vessel to be the comparison area, wherein a median value of a blood vessel diameter of the second area less than a median value of a blood vessel diameter of the first area, and calculate, using the processor, a stenosis score indicating a degree of stenosis of a blood vessel of the stenosis area compared to a blood vessel of the comparison area.

13 . The apparatus of claim 12 , wherein the processor is further configured to
detect an area comprising a blood vessel diameter decreasing in the longitudinal direction of the target blood vessel as the comparison area in an area different from the stenosis area of the blood vessel area.

14 . The apparatus of claim 12 , wherein the processor is further configured to
determine a partial centerline based on a partial reference line corresponding to the stenosis area of a reference line of the target blood vessel, and determine an area corresponding to the determined partial centerline to be the comparison area.

15 . The apparatus of claim 12 , wherein the processor is further configured to
determine the comparison area corresponding to a normal blood vessel adjacent to the stenosed portion of the stenosis area of the blood vessel area.

16 . The apparatus of claim 12 , wherein the processor is further configured to
extract blood vessel diameters corresponding to the stenosis area and the comparison area from the distribut

Assignees

• Catholic Univ Korea Ind Academic Cooperation Foundation

Inventors

• Kim Namkug
• Shin Keewon
• Kim Dae-Won

Classifications

• G06V10/20

  Image preprocessing · CPC title

• G06V40/14

  Vascular patterns · CPC title

• G06T7/11

  Region-based segmentation · CPC title

• A61B6/504

  for diagnosis of blood vessels, e.g. by angiography · CPC title

• G06N20/00

  Machine learning · CPC title