Using machine learning algorithms based on patient admission laboratory parameters to predict adverse outcomes in COVID-19 patients

• DOI: 10.1016/j.heliyon.2024.e29981
• Publication Date: 2024-04-21T17:22:30Z
• Authors: Yuchen Fu; Xuejing Xu; Juan Du; Taihong Huang; Jiping Shi; Guanghao Song; Qing Gu; Han Shen; Sen Wang
• Keywords: 2019-20 coronavirus outbreak

Amidst the global COVID-19 pandemic, urgent need for timely and precise patient prognosis assessment underscores significance of leveraging machine learning techniques. In this study, we present a novel predictive model centered on routine clinical laboratory test data to swiftly forecast survival outcomes upon admission. Our integrates feature selection algorithms binary classification algorithms, optimizing algorithmic through meticulous parameter control. Notably, developed an algorithm coupling Lasso SVM methodologies, achieving remarkable area under ROC curve 0.9277 with use merely 8 parameters collected primary contribution lies in utilization straightforward prognostication, circumventing processing intricacies, furnishing clinicians expeditious prognostic tool.