Health Risk Analysis Expert System for Family Caregiver of Person with Disabilities using Data Mining Techniques

Abstract

The nursing care for the family caregiver of the disabled person is an important task for long-term care, since the caring people with disabilities is the difficult and hard task. In this paper, the Health Risk Analysis System or HRAS is introduced which is the new expert system for identifying the health risk level in three aspects including mental, physical, and social health aspects, and provides the intervention according to the health risk level of each aspect as well. The HRAS is the client-server system. The HRAS client proceeds on web-based application to collect health data via online questionnaire and shows the analysis results. The collected health data are transmitted to the server to analysis and to assess the health risk level by using the proposed classifier model named Risk Analysis Classifier or RAC. The classification algorithm and rule-based classifier are used to build the RAC. The RAC is evaluated using k-fold cross validation and the experts with annotated health data and unseen data. The evaluation results showed that Neural Network performs the best performance overall which it achieves the accuracy above 90% in all health data sets. Thus, the Neural Network is the most suitable classifier for this work. In addition, the HRAS has been deployed and collected the user experience via the formal survey. These survey results demonstrated that the system provides high accuracy assessment and very utilization in several aspects.

Artificial Intelligence Support for Informal Patient Caregivers: A Systematic Review

“…Meanwhile, other research has introduced advanced systems for evaluating caregivers' health risks across mental, physical, and social domains, offering tailored interventions. Notably, a study achieved a maximum accuracy of 95.36% and a precision of 95.95% in the social domain when using a neural network [55].…”

“…The studies in the reviewed articles were conducted from 2012 to 2023 across various countries, including the Netherlands [53], Thailand [55], Italy [47], Canada [56], Germany [50,52], Ireland [48,49], Korea [51], and United States [57].…”

“…Eight of these studies primarily employed machine learning models [47][48][49][50][51][52]55,57], one utilized a chatbot [56], and another incorporated model-based reasoning [53]. Additionally, one study developed a novel health risk analysis system, based on a risk analysis classifier (RAC) and using the SCUT algorithm [55].…”

“…Eight of these studies primarily employed machine learning models [47][48][49][50][51][52]55,57], one utilized a chatbot [56], and another incorporated model-based reasoning [53]. Additionally, one study developed a novel health risk analysis system, based on a risk analysis classifier (RAC) and using the SCUT algorithm [55]. Regarding validation methods, four studies [47,49,50,57] implemented Hold-Out validation, one solely used 10-fold CV [48], and another deployed both 2-fold CV and 10-fold CV [52].…”

“…Studies show that many family caregivers face health issues due to insufficient exercise, neglecting annual health check-ups, and experiencing sleep disturbances [55]. Utilizing web-derived health data, Suksawatchon et al [55] developed the health risk analysis system (HRAS).…”