Design and Implementation of a Prototype for Frailty Diseases Based on Decision Support System

Abstract

An older people's disease known as frailty occurs when an elderly individual exhibits heightened susceptibility to little stimuli. An individual who is feeble experiences limited ability to grasp, an unintended reduction in weight, a feeling of happiness, a sluggish gait, and trouble standing. The detection of frailty phenomenology is one of the greatest developments in geriatric medicine in the past 20 years. Frailty develops into an impairment if left unchecked. Clinical decision support systems are used to help professionals, or physicians, make more informed and complex decisions. Thus, in order to assist the board in providing healthcare effortlessly in controlling the frailty conditions and, usually, to identify the class of acquiring a medical condition (frailty) and offer an appropriate procedure therapy based on what can be placed to the perfect use, an appropriate support system has been set up using doctors' expertise and information mining extraction structure. In order to categorize this condition and analyze their respective efficacy and rectification rates, we suggest using the Decision Trees, C4.5, ID3, Random Forests (RF), Logistic Regression (LR), and CART algorithms. With a decrease in undertriaging elderly patients who are seriously ill, predictive machine learning assessment demonstrated superior discriminating capacity for predicting medical results and attitudes. The present research identifies an effective combination of six characteristics—unintentional reduction in weight, sluggish weak strength or tiredness, frequent chair current stands, and age—that assist in predicting death with precision and F1 score using predictive machine learning techniques. In order to obtain consistently high accuracy throughout the characteristics of frailty illness detection, numerous algorithms based on machine learning are being examined. Five cases of rigorous testing validate the excellent predictive accuracy of the suggested framework. This research would hasten the process of making decisions in healthcare organizations so that targeted therapies can be administered promptly and precisely.