[Machine learning-based method for interpreting the guidelines of the diagnosis and treatment of COVID-19]

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Public Health

Background:

• The COVID-19 pandemic necessitated rapid development and frequent updates of treatment guidelines.
• Distinguishing key changes across multiple versions of the "Guidelines for the Diagnosis and Treatment of COVID-19" proved challenging for clinicians and the public.
• Effective communication of evolving medical protocols is crucial during public health emergencies.

Purpose of the Study:

• To develop a computer-aided intelligent analysis method using machine learning.
• To automatically analyze similarities and differences between different versions of COVID-19 treatment guidelines.
• To present the focus of new guideline versions to clinicians and simplify understanding for the public.

Main Methods:

• Utilized machine learning, specifically unsupervised learning, for topic prediction and matching.
• Trained the model on previous versions of the "Guidelines for the Diagnosis and Treatment of COVID-19".
• Developed a method for computer-aided intelligent analysis of treatment plan texts.

Main Results:

• Achieved 100% accuracy in topic prediction and matching for new guideline versions.
• Demonstrated the ability of the method to automatically analyze similarities and differences in treatment plans.
• Successfully enabled intelligent computer interpretation of diagnosis and treatment plans.

Conclusions:

• The developed machine learning method effectively identifies key updates in COVID-19 treatment guidelines.
• This approach enhances the ability of healthcare professionals to quickly grasp new information.
• The system facilitates better public comprehension of complex medical guidance during health crises.