Enhancing Clinical Decision Support System : Risk Stratification of Hepatological Disorders by Advanced Machine Learning Models
Year: 2026
Type: article
Abstract: Artificial Intelligence (AI) is reshaping healthcare services reflecting the increasing significance of data science for disease prediction. The integration of AI into diagnostic processes helps to optimize many processes in healthcare management. This research emphasis to enhance physician productivity in disease prediction, particularly concerning chronic hepatological disorders, to increase care management. It proposes a comprehensive framework for the incorporation of ensemble machine learning models into clinical practice. Furthermore, healthcare operations and strategic management data, informed by statistical analysis and AI can facilitate the identification of critical success factors for the healthcare management system through the proposed digital Healthcare Transformation. The methodology is driven by patient data derived taken from health systems at the University of California, Irvine (UCI), and encompasses the mapping of management diagnostics workflows. To refine the process of illness diagnosis, the proposed approach formulates a clinical decision-making algorithm leveraging contemporary AI technologies. The objectives of this study encompass mitigating physician burnout with streamlining care administration and augmenting diagnostic accuracy. This research highlights the importance of enhancing diagnostic accuracy and maximizing physician effectiveness for the treatment of liver diseases. Enhanced model performance efficiency, as indicated by KPIs is achieved through the implementation of a state of the art ensemble machine learning methodology. Our approach also emphasis on the comparative predictive strengths of individual Artificial Intelligence algorithms including Gradient Boosting (GB), Random Forest Classifier (RF) and Artificial Neural Network (ANN). To improve the accuracy and reliability of Hepatological Disorder predictions, the proposed Voting Ensemble Classifier integrates the outputs of these algorithms, thus addressing the limitations of using in a single model which may be vulnerable to data noise or overfitting.
Author: Chandan Pan
Institution: Indian Institute of Technology Kharagpur
Language: English
Subfield: Health Information Management
Field: Health Professions
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W7147197546
