Method, apparatus and computer program for measuring and analysis of blood pressure using photoplethysmography

Patent metadata

Field	Value
Publication number	US-2020375550-A1
Application number	US-202016885665-A
Country	US
Kind code	A1
Filing date	May 28, 2020
Priority date	May 30, 2019
Publication date	Dec 3, 2020
Grant date	—

Abstract

A method for measuring and analyzing blood pressure using PPG includes receiving, by a computer, a PPG signal from a finger of a subject, dividing, by the computer, a normalization pulse wave signal derived from the received PPG signal into one or more predetermined windows, extracting, by the computer, a maximum lower amplitude value from one of the respective divided windows, extracting, by the computer, a target feature pattern from the extracted maximum lower amplitude value, deriving, by the computer, a first target unique vector and a second target unique vector with respect to the target feature pattern, using a linear discriminant analysis (LDA) algorithm to display the first target unique vector and the second target unique vector of the target feature pattern on 2-dimensional (2D) graph, and providing, by the computer, a blood pressure state of the subject, using the 2D graph.

First claim

What is claimed is:

1 . A method for measuring and analyzing blood pressure using photoplethysmography (PPG), the method comprising:
receiving, by a computer, a PPG signal from a finger of a subject; dividing, by the computer, a normalization pulse wave signal derived from the received PPG signal into one or more predetermined windows, wherein the windows are divided based on a predetermined window region range; extracting, by the computer, a maximum lower amplitude value from one of the respective divided windows; extracting, by the computer, a target feature pattern from the extracted maximum lower amplitude value; deriving, by the computer, a first target unique vector and a second target unique vector with respect to the target feature pattern, using a linear discriminant analysis (LDA) algorithm to display the first target unique vector and the second target unique vector of the target feature pattern on 2-dimensional (2D) graph; and providing, by the computer, a blood pressure state of the subject, using the 2D graph.

2 . The method of claim 1 , wherein the receiving of the PPG signal includes:
receiving, by the computer, the PPG signal from an index finger of a right hand and an index finger of a left hand of the subject, and wherein the PPG signal includes a PPG signal of the index finger of the right hand and a PPG signal of the index finger of the left hand.

3 . The method of claim 1 , wherein, in the dividing, the normalization pulse wave signal is a value of a normalized pulse wave signal by using a value by overall performing square root on a value from adding values obtained by squaring one or more pulse wave signals derived from the PPG signal received by the computer.

4 . The method of claim 1 , wherein the first target unique vector and the second target unique vector are each unique vector value of the subject having a first unique vector in a case of the first target unique vector and a second unique vector in a case of the second target unique vector as axes,
wherein the first unique vector and the second unique vector are derived such that each feature pattern of the maximum lower amplitude value derived depending on each pulse wave signal of a normal group, a high blood pressure risk group, and a high blood pressure group is classified into the normal group, the high blood pressure risk group, and the high blood pressure group by using the LDA algorithm, wherein the first unique vector is a first unique vector that occupies a largest classification weight of the feature pattern, wherein the second unique vector is a second unique vector that occupies a second largest classification weight of the feature pattern, and wherein the display includes: displaying the first target unique vector and the second target unique vector of the target feature pattern with the 2D graph, using the first unique vector and the second unique vector, which are derived to be classified into the normal group, the high blood pressure risk group, and the high blood pressure group, as axes.

5 . The method of claim 1 , wherein the providing of the blood pressure state includes:
providing the blood pressure state as a blood pressure state of any one of a normal group, a high blood pressure risk group, and a high blood pressure group.

6 . The method of claim 5 , wherein the providing of the blood pressure state includes:
providing an analysis result considering a factor other than the blood pressure with respect to an overlapping section together when two or more of the normal group, the high blood pressure risk group, and the high blood pressure group overlap with one another.

7 . The method of claim 6 , wherein the factor other than the blood pressure includes at least one of age, weight, workout, whether to take blood pressure medication, elasticity of a blood vessel, calciumation of a blood vessel, or decreased elasticity on capillaries.

8 . The method of claim 6 , wherein the providing of the blood pressure state includes:
when elasticity of a blood vessel is high, providing the analysis result together in consideration of a case where blood pressure of the subject is classified into the high blood pressure risk group even though the blood pressure of the subject should be classified as the high blood pressure group or a case where the blood pressure of the subject is classified into the normal group even though the blood pressure of the subject should be classified as the high blood pressure risk group, and when the elasticity of the blood vessel is low, providing the analysis result together in consideration of a case where the blood pressure of the subject is classified into the high blood pressure group even though the blood pressure of the subject should be classified as the high blood pressure risk group or a case where the blood pressure of the subject is classified into the high blood pressure risk group even though the blood pressure of the subject should be classified as the normal group.

9 . A computer program for measuring and analyzing blood pressure using PPG that is stored in a medium to execute the method of claim 1 in combination with a computer that is hardware.

10 . A computer apparatus for measuring and analyzing blood pressure using PPG, wherein the computer apparatus:
receives a PPG measurement signal from a finger of a subject; divides a normalization pulse wave signal derived from the received PPG measurement signal into one or more predetermined windows, wherein the windows are divided based on a predetermined window region range; extracts a maximum lower amplitude value from one of the respective divided windows; extracts a target feature pattern from the extracted maximum lower amplitude value; derives a first target unique vector and a second target unique vector with respect to the target feature pattern, using a LDA algorithm to display the first target unique vector and the second target unique vector of the target feature pattern on a 2D graph; and provides a blood pressure state, using the 2D graph.

Assignees

• Univ Chosun Iacf

Inventors

• Shin Young Suk

Classifications

• A61B5/7246

  using correlation, e.g. template matching or determination of similarity · CPC title

• A61B5/0261

  using optical means, e.g. infrared light · CPC title

• A61B5/0059

  using light, e.g. diagnosis by transillumination, diascopy, fluorescence (photoacoustic A61B5/0093; optical measurement of heart rate A61B5/02416; optical measurement of blood flow A61B5/0261; optical measurement of analytes A61B5/1455) · CPC title

• A61B5/7275

  Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor · CPC title

• A61B5/6826

  Finger · CPC title