Principal Component Analysis Usage in Biomedical Engineering to Aid at Diagnosing Pathologies

Abstract

-This paper aims to present a review of the many uses of the statistical method of Principal Component Analysis (PCA) in the Engineering Biomedical field, aimed specially in those where PCA was used as a tool to diagnose pathologies in the last 5 years. An exploratory study was made through the use of bibliometrics, narrowing down the initial search to a final portfolio of 26 papers, providing the latest and state-of-the-art researches on the desired field of study. It was found that PCA has been used in a wide spectrum of areas with significant results, and all around the world. The main use is to reduce the dimensionality of the data to a few principal variables which can explain most of the variance present in the original data. There were studies which reduced from 14 to 50 variables into 1 to 6 principal components, while retaining in average 80% of the variance, and others reduced from 51 to 140 variables into as low as 2 components, keeping 68% to 99% of the variance.