Dimensionality Reduction for IoMT Devices Using PCA
Rajiv Pandey, Radhika Awasthi, Archana Sahai, Pratibha Maurya
Abstract
Data power our world. Data give us instantaneous intelligence and fundamentally help find answer to a myriad of problems. Currently, the healthcare sector generates over 30% of the total amount of data worldwide. Because of its efficiency in collecting, analyzing, and transmitting health data, Internet of Medical Things (IoMT) has the potential to alter the medical world with the help of developing technologies. Yet, high-dimensionality and linear correlations can occasionally be the downfall of data. In this chapter, we propose an analysis that uses Principal Component Analysis (PCA), one of the most widely used dimensionality reduction techniques, to help resolve the dimensionality problem of data. We assess the significance of utilizing PCA in reducing the dimensions of the dataset used in an IoMT-enabled system, coupling our research with a previously proposed framework called “Prenatal Healthcare System of Remote Mother and Fetal Surveillance via IoMT.” The prenatal device increases the likelihood of a safe and healthy birth while lowering pregnancy risks. A fetus’s survival depends on routine health assessments, which are both advantageous and essential. The dataset used in the experimentation is made up of the crucial prenatal device parameters for a fetus, which may aid in providing medical experts with real-time health updates. We implement PCA to highlight variation and bring out significant patterns in the dataset in order to accurately forecast outcomes.