AI-powered medical decision support system that does not require internet access for use in rural and remote areas

The work is devoted to assessing the possibility of using a large open-source language model running on a local computer as an additional source of information for a doctor in matters of diagnosis and treatment of diseases. Preloaded texts of clinical guide lines (arterial hypertension and diabetes mellitus) were used. Testing showed acceptable accuracy of neural network responses. It is assumed that the created support system for medical decision-making can be used by medical workers in rural areas without access to the Internet.