Combined Multimodal Clinical And Radiological Model For Classifying Soft Tissue Tumors

Soft tissue tumors (STTs) represent a heterogeneous group of mesenchymal neoplasms ranging from benign lesions to highly aggressive sarcomas. Accurate preoperative classification is essential for guiding biopsy planning, surgical strategy, and systemic therapy. However, conventional diagnostic workflows—based on clinical assessment and imaging interpretation—often face limitations due to overlapping morphological features across tumor subtypes. A combined multimodal clinical and radiological model aims to improve diagnostic performance by integrating structured clinical data (e.g., age, sex, tumor location, growth rate, symptom duration, laboratory parameters) with advanced radiological features derived from imaging modalities such as MRI, CT, and occasionally PET. Radiological inputs may include conventional imaging characteristics (size, margins, signal intensity, contrast enhancement patterns) as well as quantitative radiomic features capturing tumor texture, shape, and heterogeneity.