MULTIMODAL CLINICAL-IMAGING DATASET FOR RISK STRATIFICATION OF SOLITARY PULMONARY LESIONS USING ARTIFICIAL INTELLIGENCE

Modern medical imaging techniques generate substantial volumes of clinical, morphological, and quantitative information on solitary pulmonary lesions. However, patients’ clinical characteristics, multislice computed tomography (MSCT) findings, and metabolic parameters obtained by positron emission tomography/computed tomography (PET/CT) are often assessed separately. This approach may limit the potential for comprehensive risk stratification. Integration of multimodal data using artificial intelligence (AI) and machine-learning methods represents a promising approach for improving the objectivity of malignancy risk assessment and supporting clinical decision-making.