MRI-Based Radiomics Features for Prediction of Pathological Deterioration Upgrading in Rectal Tumor.

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• Accurate prediction of pathological upgrading in rectal tumors is crucial for treatment planning.
• Current diagnostic methods may not fully capture the heterogeneity of rectal tumors.

Purpose of the Study:

• To develop and validate a diagnostic model using Magnetic Resonance Imaging (MRI) for predicting pathological deterioration upgrading in rectal tumors.
• To assess the performance of radiomics, clinical features, and a combined model in predicting pathological upgrading.

Main Methods:

• Retrospective study of 158 rectal tumor patients.
• Extraction of radiomics features from T2-weighted MRI images.
• Development of radiomics, clinical, and combined radiomics-clinical models.
• Evaluation using Receiver Operating Characteristic (ROC) analysis and Area Under the ROC Curve (AUC).

Main Results:

• The radiomics model showed good predictive performance (AUC 0.863 training, 0.861 validation).
• The clinical model achieved AUCs of 0.669 (training) and 0.651 (validation).
• The combined radiomics-clinical model significantly outperformed individual models (AUC 0.932 training, 0.907 validation).

Conclusions:

• A combined radiomics-clinical model based on MRI effectively predicts pathological upgrading in rectal tumors.
• This model provides valuable insights for developing personalized treatment strategies.
• The findings support the use of advanced imaging analysis in rectal cancer management.