Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Machine Learning Assisted Postural Movement Recognition using Photoplethysmography(PPG)

Robbie Maccay, Roshan Weerasekera

TL;DR

Various machine learning approaches were used for classification, and the Artificial Neural Network was found to be the best classifier, with a testing accuracy of 85.2\% and an F1 score of 78\% from experimental results.

Abstract

With the growing percentage of elderly people and care home admissions, there is an urgent need for the development of fall detection and fall prevention technologies. This work presents, for the first time, the use of machine learning techniques to recognize postural movements exclusively from Photoplethysmography (PPG) data. To achieve this goal, a device was developed for reading the PPG signal, segmenting the PPG signals into individual pulses, extracting pulse morphology and homeostatic characteristic features, and evaluating different ML algorithms. Investigations into different postural movements (stationary, sitting to standing, and lying to standing) were performed by 11 participants. The results of these investigations provided insight into the differences in homeostasis after the movements in the PPG signal. Various machine learning approaches were used for classification, and the Artificial Neural Network (ANN) was found to be the best classifier, with a testing accuracy of 85.2\% and an F1 score of 78\% from experimental results.

Machine Learning Assisted Postural Movement Recognition using Photoplethysmography(PPG)

TL;DR

Various machine learning approaches were used for classification, and the Artificial Neural Network was found to be the best classifier, with a testing accuracy of 85.2\% and an F1 score of 78\% from experimental results.

Abstract

With the growing percentage of elderly people and care home admissions, there is an urgent need for the development of fall detection and fall prevention technologies. This work presents, for the first time, the use of machine learning techniques to recognize postural movements exclusively from Photoplethysmography (PPG) data. To achieve this goal, a device was developed for reading the PPG signal, segmenting the PPG signals into individual pulses, extracting pulse morphology and homeostatic characteristic features, and evaluating different ML algorithms. Investigations into different postural movements (stationary, sitting to standing, and lying to standing) were performed by 11 participants. The results of these investigations provided insight into the differences in homeostasis after the movements in the PPG signal. Various machine learning approaches were used for classification, and the Artificial Neural Network (ANN) was found to be the best classifier, with a testing accuracy of 85.2\% and an F1 score of 78\% from experimental results.

Paper Structure

This paper contains 20 sections, 2 equations, 14 figures, 1 table.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. Photoplethysmography
  4. Machine Learning
  5. Discriminant Analysis (DA):
  6. Naive Bayes (NB):
  7. Support Vector Machines (SVM):
  8. Decision tree (DT):
  9. K-Nearest Neighbours (KNN):
  10. Artificial Neural Network (ANN):
  11. Human Activity Recognition
  12. Method ana Analysis
  13. Hardware Design
  14. Software Design
  15. Data Collection
  16. ...and 5 more sections

Figures (14)

  • Figure 1: PPG based postural movement detection methodology
  • Figure 2: Principles of the PPG signal ref15
  • Figure 2: Machine learning algorithm evaluations with accuracy scores for each model and F1 scores for each class. S = Stationary, SS = Sitting-to-standing and LS = Lying-to-standing.
  • Figure 3: PPG pulse points of interest and features
  • Figure 4: Examples of distortions in the PPG signal due to motion artifacts, baseline wandering and hypoperfusion ref15
  • ...and 9 more figures