Machine Learning Approach to Predicting COVID-19 Disease Severity Based on Clinical Blood Test Data: Statistical...

Area of Science:

• Medical Informatics
• Computational Biology
• Epidemiology

Background:

• Accurate prediction of COVID-19 severity is vital for patient care and resource management, especially in resource-limited settings.
• Pre-existing patient comorbidities significantly influence disease progression and can aid in severity prediction.
• Routine peripheral blood sample analysis offers a readily available data source for predicting clinical outcomes.

Purpose of the Study:

• To investigate the utility of peripheral blood data for predicting clinical outcomes in COVID-19 patients.
• To develop and validate machine learning models for automated disease severity profiling.

Main Methods:

• Utilized statistical comparison and correlation methods on COVID-19 patient datasets with known outcomes.
• Applied various machine learning algorithms including decision tree, random forest, gradient boosting, support vector machine, k-nearest neighbor, and deep learning.
• Integrated clinical data from peripheral blood samples for predictive modeling.

Main Results:

• Identified specific clinical parameters in blood samples that effectively differentiate healthy individuals from COVID-19 patients.
• Demonstrated the value of these blood parameters in predicting the later severity of COVID-19 symptoms.
• Developed analytical methods achieving >90% accuracy and precision in disease severity prediction.

Conclusions:

• Developed methodologies for analyzing routine clinical data to accurately predict COVID-19 patient outcomes.
• Standard hospital laboratory blood tests can identify high-risk COVID-19 patients.
• This approach enables optimized hospital resource allocation for effective COVID-19 treatment.