Scholarly Publication

Permanent URI for this collection

https://kwasuspace.kwasu.edu.ng/handle/123456789/48

Browse

Browsing Scholarly Publication by Author "Babatunde Roneke Seyi"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    A Clinical Decision Support System for Predicting Psychotic Disorder using Random Forest Algorithm
    (Published by Faculty of Engineering and Technology, Adeleke University. Ede. Osun State, 2023-05-19) Babatunde Roneke Seyi
    Psychotic disorders exert a profound effect on one's perception of reality, cognitive abilities, emotions, and conduct. By precisely pinpointing individuals at risk or in the initial phases of a psychotic disorder, healthcare providers can promptly implement interventions, employ suitable treatment approaches, and offer essential support. This proactive approach holds the potential to impede the advancement of the disorder and mitigate its long-term impact on the individual's well-being. Earlier automated predictive systems are limited in accuracy of prediction, ease of use and complexity of computation. This work implements a Clinical Decision Support System (CDSS) utilizing the Random Forest algorithm. A dataset consisting of medical records from 500 carefully selected psychotic patients at Yaba Psychiatric Hospital in Lagos, Nigeria, spanning a period of five years (January 2010 to December 2014) was obtained. The dataset provides crucial insights into the demographic and clinical characteristics of the patients, enabling the Random Forest algorithm to capture relevant patterns and relationships associated with psychotic disorders. Following model training and evaluation, employing a range of evaluation metrics such as cross-validation, F1 score, recall score, and precision score, the predictive model achieves an impressive accuracy rate of approximately 95%. The implications of this research are profound. By harnessing the power of machine learning and the Random Forest algorithm, the CDSS holds great potential in significantly enhancing psychiatric diagnoses. The accuracy attained by the predictive model showcases its reliability and effectiveness as a decision support tool for healthcare professionals. This technology has the capacity to expedite diagnosis, enabling timely interventions and personalized treatment plans for patients with psychotic disorders.