CT-based multi-phase Radiomic models for differentiating clear cell renal cell carcinoma

Area of Science:

• Radiology
• Oncology
• Medical Imaging Analysis

Background:

• Differentiating clear cell renal cell carcinoma (ccRCC) from non-ccRCC is crucial for treatment.
• CT-based texture and non-texture features show potential for improved renal tumor classification.

Purpose of the Study:

• To compare the diagnostic value of CT texture and non-texture features for differentiating ccRCC from non-ccRCC.
• To evaluate the efficacy of combined feature models in ccRCC diagnosis.

Main Methods:

• 197 pathologically proven renal tumors were analyzed.
• 43 non-texture and 296 texture features were extracted across four CT phases (NCP, CMP, NP, EP).
• Three models were developed: non-texture features only, texture features only (texture scores), and combined features.

Main Results:

• All three models demonstrated good discrimination between ccRCC and non-ccRCC.
• The combined model (Model 3) achieved the highest Area Under the Curve (AUC) values, ranging from 0.864 to 0.900.
• The differences in AUC between all models were statistically significant (p < 0.001).

Conclusions:

• Combining non-texture and texture features significantly enhances the diagnostic performance for ccRCC.
• A combined diagnostic model provides a reliable basis for clinical treatment decisions in renal cell carcinoma.