Fact Finding Instructor-based Clustering Technique for BP Estimation using Human Speech Signals | JoVE Visualize

Area of Science:

Biomedical Engineering
Signal Processing
Machine Learning

Background:

Accurate blood pressure (BP) monitoring is crucial for cardiovascular health, yet current estimation methods face accuracy challenges.
High and low BP pose significant health risks, necessitating reliable and accessible monitoring techniques.
Speech signals contain physiological information that can potentially be leveraged for non-invasive BP estimation.

Purpose of the Study:

To develop an efficient and accurate blood pressure (BP) estimation technique using speech signals.
To introduce a novel Fact Finding Instructor (FFI) optimization algorithm integrated with k-means clustering for BP classification.
To evaluate the performance of the proposed FFI-based method against existing techniques.

Main Methods:

Speech signals were collected, pre-processed to remove noise, and feature extraction was performed to generate feature vectors.
A hybrid clustering approach combining the proposed Fact Finding Instructor (FFI) optimization algorithm and k-means clustering was employed.
Blood pressure was classified into three categories: Low BP, Normal, and High BP, based on the clustering results.

Main Results:

The FFI optimization algorithm demonstrated a fast convergence rate and accurate clustering of patient BP categories from speech features.
The combined FFI and k-means clustering method achieved significant performance improvements over the Teaching Learning Based Optimization (TLBO) clustering technique.
Evaluation metrics including Davies Bouldin, Homogeneity, Completeness, Jacquard Similarity, Silhouette, and Dunn's Index showed improvement rates up to 0.98.

Conclusions:

The proposed Fact Finding Instructor (FFI) based clustering technique offers an efficient and accurate method for blood pressure estimation using speech signals.
This speech-based approach provides a promising non-invasive alternative for regular BP monitoring and healthcare maintenance.
The FFI algorithm's effectiveness in feature extraction and clustering highlights its potential for various physiological signal analysis applications.