This retrospective study combine radiomics and deep learning models to predict malignancy in Bosniak II-III cystic renal masses, aiming for improving preoperative risk assessment and reducing unnecessary surgery for benign lesions and avoiding delayed treatment of malignant disease. The main question it aims to answer is: * How to specially predict malignancy in Bosniak II-III cystic renal masses？ The investigators retrospectively included patients diagnosed with Bosniak II-III cystic renal masses based on preoperative contrast-enhanced CT.
Patients diagnosed with Bosniak II-III cystic renal masses based on preoperative contrast-enhanced CT were included. The exclusion criteria were solid portion \> 25%, polycystic kidney disease, maximum diameter\<1cm, Von Hippel-Lindau syndrome, without complete CT examination or histopathology-proven CRMs, poor image quality and Bosniak I and Bosniak IV masses. Deep learning models were developed to classify renal cystic lesions as benign or malignant using multiphase CT images. The performance measure were the area under the receiver operating characteristic curve, sensitivity, specificity, and balanced accuracy.
Study Type
OBSERVATIONAL
Enrollment
223
Zhongshan Hospital Fudan University, Location: 180th Fenglin Road, Xuhui District, Shanghai, China
Shanghai, Shanghai Municipality, China
the area under the receiver operating characteristic curve
The area under the receiver operating characteristic curve measures the overall ability of a binary classifier to distinguish between positive and negative classes, with values ranging from 0.5 (random guessing) to 1.0 (perfect classification).
Time frame: preoperatively
sensitivity
Sensitivity measures the proportion of actual positive cases that are correctly identified by the model. A higher sensitivity indicates that the model is better at detecting positive cases and produces fewer false negatives.
Time frame: preoperatively
specificity
Specificity measures the proportion of actual negative cases that are correctly identified by the model. A higher specificity indicates that the model is better at correctly excluding negative cases and produces fewer false positives.
Time frame: preoperatively
balanced accuracy
Balanced accuracy is the average of sensitivity and specificity. It provides an overall measure of classification performance while giving equal importance to the positive and negative classes, making it particularly useful when the classes are imbalanced.
Time frame: preoperatively
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.