ETIOLOGICAL IDENTIFICATION OF RESPIRATORY AND ANGINA SYNDROMES USING ARTIFICIAL INTELLIGENCE BASED ON CLINICAL SYMPTOMS AND LABORATORY MARKERS

Infectious diseases presenting with respiratory and angina syndromes are widely prevalent in clinical practice, and early and accurate etiological diagnosis is crucial for selecting effective treatment strategies. This study evaluates the effectiveness of artificial intelligence (AI) technologies in etiological diagnosis based on clinical symptoms and laboratory markers. Machine learning models (random forest, gradient boosting, logistic regression) were applied for multiparametric analysis, demonstrating that integration of CRP, procalcitonin, leukocyte and lymphocyte counts, and clinical features enables high diagnostic accuracy (AUC ≥ 0.94) in differentiating bacterial and viral etiologies. The AI-based approach reduces diagnostic errors and optimizes clinical decision-making.以呼吸道综合征与咽峡炎综合征为临床表现的感染性疾病在临床实践中广为流行，早期精准的病因学诊断对于遴选有效治疗策略至关重要。本研究评估了基于临床症状与实验室检测标志物的人工智能（AI）技术在病因学诊断中的效能。本研究采用机器学习模型（随机森林（random forest）、梯度提升（gradient boosting）、逻辑回归（logistic regression））开展多参数分析，结果显示，整合C反应蛋白（CRP）、降钙素原、白细胞计数、淋巴细胞计数及临床特征，可实现区分细菌与病毒病因的高诊断准确度，其受试者工作特征曲线下面积（AUC）≥0.94。该基于AI的诊断方法可降低诊断误差，优化临床决策制定。