A Decision-Making Algorithm for the Diagnosis of Lung Diseases Based on the Integrated Analysis of Respiratory Sound Signals and Medical Images
Ochiq kirish: yoʻq
Annotatsiya
This paper presents a formalized decisionmaking algorithm for the diagnosis of lung diseases based on the integrated analysis of respiratory sound signals and medical images. The proposed approach reflects the clinical logic of the diagnostic process and is oriented toward the support of clinical decisions and the implementation of automated intelligent systems. The algorithm structures the diagnostic stages from the acquisition of clinical data and the analysis of auscultatory signals to the interpretation of medical images and the formation of a substantiated diagnostic conclusion. Particular attention is given to the integration of multimodal data and to the application of Explainable Artificial Intelligence principles. The clinical evaluation involved 102 patients with pulmonary pathology (20 with bronchial asthma, 31 with pneumonia, 44 with chronic obstructive pulmonary disease, and 7 with lung cancer). The proposed multimodal model achieved an accuracy of 95.1%, an F1 score of 0.94, and an AUC of 0.97, significantly outperforming the audio-only and image-only baselines
Hali tarjima qilinmagan