A CT-based radiomics nomogram for distinguishing between benign and malignant bone tumours

Area of Science:

• Medical Imaging
• Oncology
• Radiology

Background:

• Bone tumors present diagnostic challenges in distinguishing benign from malignant types.
• Accurate preoperative differentiation is crucial for effective treatment planning.

Purpose of the Study:

• To evaluate a novel computed tomography (CT)-based radiomics nomogram for differentiating benign and malignant bone tumors.
• To assess the diagnostic performance and clinical utility of the developed radiomics nomogram.

Main Methods:

• A radiomics nomogram was developed using features extracted from unenhanced CT images of 206 bone tumor patients.
• Least absolute shrinkage and selection operator logistic regression was employed for feature selection and model construction.
• The radiomics nomogram integrated clinical data and radiomics signatures, and its performance was compared against clinical and radiomics signature models alone.

Main Results:

• The radiomics nomogram demonstrated strong performance in both training (AUC 0.917) and validation (AUC 0.823) sets.
• Decision curve analysis and net reclassification improvement indicated superior diagnostic performance and clinical utility compared to the clinical model alone.
• The combined model achieved better diagnostic accuracy and greater clinical net benefits.

Conclusions:

• A combined radiomics nomogram integrating clinical and radiomics features serves as an effective noninvasive preoperative prediction tool.
• This nomogram assists in distinguishing benign from malignant bone tumors, thereby aiding treatment planning.
• The developed model offers improved diagnostic performance and clinical value for bone tumor management.