Intelligent support for pulmonologists based on a multi-level patient computer model

Abstract: The paper presents models, algorithms, and a software package for intelligent support of pulmonologists based on a multilevel patient computer model. The developed package includes diag-nostic and therapy selection modules integrated with a relational database and a web application featuring user role differentia-tion. Mamdani fuzzy inference algorithms were used to formal-ize expert knowledge, ensuring interpretability of decisions and robustness to incomplete initial data. Testing on real clinical data of patients with bronchial asthma confirmed the correct op-eration of the software package: the system demonstrated agree-ment between its results and expert medical opinions and the ability to generate reliable recommendations even when some input parameters were missing. The practical significance of the work lies in reducing data analysis time, standardizing the diag-nostic and therapeutic process, and increasing the transparency in healthcare organizations.

Keywords: database, web application, therapy, diagnos-tics, mamdani algorithm, fuzzy logic, multi-level patient model, pulmonology, intelligent decision support