Renal cell carcinoma
Conditions
Interventions
Index test:A Deep Learning-Based Diagnostic Model for Benign and Malignant Renal Tumors Using Ultrasound Video Images
Sponsors
Peking University First Hospital
Eligibility
Sex/Gender
All
Age
18 Years to No maximum
Inclusion criteria
Inclusion criteria: Retrospective cohort: 1. Age >= 18 years. 2. Underwent renal ultrasound examination at our hospital and had a renal mass detected. 3. Possesses a complete ultrasound diagnostic report, with at least three static images from different cross-sectional views stored in the ultrasound department's information system. Prospective cohort: 1. Meets criteria 1–3 of the retrospective cohort. 2. A dynamic ultrasound video of more than 10 seconds is available, containing complete information about the renal tumor. 3. Clinically planned to undergo percutaneous biopsy or surgical resection for the renal lesion. 4. The patient understands the study details and voluntarily signs a written informed consent form.
Exclusion criteria
Exclusion criteria: 1. Ultrasound images/videos are of excessively poor quality, making reliable assessment impossible. 2. Key information in clinical or pathological data is missing. 3. Pregnant and breastfeeding women.
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| The diagnostic accuracy of the model in distinguishing between benign and malignant renal masses, including the AUC, sensitivity, specificity, positive predictive value, and negative predictive value.;Classification Accuracy of the Model for Different Pathological Subtypes;Accuracy of the model in predicting pTNM staging and histological grading (ISUP grading);Postoperative Survival Time;Comparison Metrics of Diagnostic Performance Between the Model and Physicians of Different Experience Levels; | — |
Countries
China
Contacts
Public ContactLi Xuesong
Peking University First Hospital
Outcome results
None listed