Related Experiment Video
Updated: Jun 25, 2025
12:24
A Three-Dimensional Spheroid Model to Investigate the Tumor-Stromal Interaction in Hepatocellular Carcinoma
Published on: September 30, 2021
5.2K
A Machine Learning-Based Mortality Prediction Model for Patients with Chronic Hepatitis C Infection: An Exploratory
Abdullah M Al Alawi1,2, Halima H Al Shuaili3, Khalid Al-Naamani3
1Department of Medicine, Sultan Qaboos University Hospital, Muscat 123, Oman.
Journal of Clinical Medicine
|May 25, 2024
Summary
Machine learning models accurately predict mortality risk in chronic hepatitis C (HCV) patients. Key factors like hemoglobin and comorbidities inform survival outcomes, aiding personalized treatment strategies.
Area of Science:
- Hepatology and viral gastroenterology
- Medical informatics and machine learning
- Public health and epidemiology
Background:
- Chronic hepatitis C virus (HCV) infection poses significant global health challenges, impacting morbidity and mortality.
- Effective treatment of HCV cirrhosis can normalize mortality rates, highlighting the need for precise risk stratification.
- Predictive modeling for HCV patient mortality is crucial for optimizing clinical management.
Purpose of the Study:
- To develop and evaluate machine learning models for predicting mortality in chronic HCV patients.
- To identify key clinical and demographic factors associated with mortality risk in this cohort.
- To enhance risk assessment and inform personalized treatment strategies for chronic HCV.
Main Methods:
- Analysis of a cohort of 702 chronic HCV patients from Sultan Qaboos University Hospital (2009-2017).
- Data pre-processing, feature selection (SelectKBest), and training of machine learning algorithms (logistic regression, random forest, gradient boosting, SVM).
- Performance evaluation using 5-fold cross-validation and metrics including Area Under the Curve (AUC).
Main Results:
- Survival probabilities at 12, 36, and 120 months were 90.0%, 84.0%, and 73.0%, respectively.
- Logistic regression achieved a high AUC of 0.929, demonstrating strong predictive power.
- Identified predictors of mortality included hemoglobin, ALT, comorbidities, HCV genotype, coinfections, follow-up duration, and treatment response.
Conclusions:
- Machine learning models effectively predict mortality in chronic HCV patients.
- Key factors influencing mortality risk were identified, enabling better patient stratification.
- These models offer valuable insights for clinical decision-making and improving patient outcomes in HCV management.
