Cuff-Less BP Stratification based on Bio-Signals processing using Machine Learning : An Investigative Study

Abstract

Cuff-Less Blood Pressure Stratification using Signal Processing with Machine Learning has gained immense attraction in the past decade among the research community. Blood Pressure, one of the most vital parameter of the human body representing overall well being of an individual. Most of the cardiovascular and cerebrovascular diseases (CCVD) including Hypertension are highly correlated to Blood Pressure. Existing BP measurement approaches are highly inconvenient and intermittent and do not allow continuous measurement of BP. Continuous BP measurement could prove to be significant indicator to most of the medicinal conditions and will lead to breakthrough achievement in the field of medical science. Cuff-Less Blood Pressure estimation hopefully can enable continuous blood pressure measurements in the time to come. A Plethora of methods for Cuff-Less BP Stratification have been experimented out by using Vital Bio Signals such PTT based, nPTT based, Machine Learning and Deep Learning based. Most of these methods leading to satisfactory beliefs that Cuff-Less BP estimation could be possible to the utmost accuracy for the diagnosis of most of the CCVD diseases as well as to monitor the overall well being of humans. However, most of the approaches still needs improvements, needs to be tested on a larger population with varying demographic features and real time application. This paper presents an investigative study of existing Cuff-Less BP Estimation approaches and discusses the merits and opportunities for improvements of the Cuff-Less BP Estimation methods.