Related Experiment Video
Updated: May 5, 2026
09:31
In Vivo, Percutaneous, Needle Based, Optical Coherence Tomography of Renal Masses
Published on: March 30, 2015
8.9K
Computed Tomography-Based Radiomics Diagnostic Model for Fat-Poor Small Renal Tumor Subtypes
Seokhwan Bang1, Heehwan Wang2, Hoyoung Bae3
1Department of Urology, Seoul St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea.
Diagnostics (Basel, Switzerland)
|June 13, 2025
Summary
Machine learning using CT radiomics effectively classifies renal tumor subtypes, including clear cell RCC and angiomyolipoma. This AI approach aids personalized diagnosis and treatment planning in renal oncology.
Area of Science:
- Oncology
- Radiology
- Artificial Intelligence
- Medical Imaging
Background:
- Differentiating fat-poor small renal masses is challenging with conventional imaging due to overlapping radiologic features.
- Accurate classification of renal tumor subtypes is crucial for effective treatment planning.
Purpose of the Study:
- To develop a machine learning (ML) model using CT-derived radiomic features for classifying five common renal tumor subtypes.
- The subtypes include clear cell renal cell carcinoma (ccRCC), papillary RCC (pRCC), chromophobe RCC (chRCC), angiomyolipoma (AML), and oncocytoma.
Main Methods:
- Retrospective analysis of 499 patients with pathologically confirmed renal tumors who underwent contrast-enhanced CT and nephrectomy.
- Extraction and analysis of radiomic features from 1548 multi-phase CT scans, focusing on fat-poor tumors.
- Evaluation of five ML classifiers: Linear SVM, Rbf SVM, Random Forest, and XGBoost.
Main Results:
- XGBoost demonstrated the best classification performance.
- Achieved an average AU-PRC of 0.757 (SE=0.033) and a renal angiomyolipoma-specific AU-ROC of 0.824 (SE=0.023).
- Performance surpassed other single-phase CT radiomic feature-based ML models.
Conclusions:
- Radiomics-based ML is effective for classifying renal tumor subtypes, showcasing AI's potential in medical imaging.
- Single-phase CT and optimized features offer valuable insights for precision medicine in renal oncology.
- These methods can support personalized diagnosis and treatment planning.
