Whole-slide Image and CT Radiomics Based Deep Learning System for Prognostication Prediction in Upper Tract Urothelial Carcinoma

Mingzhao Xiao

Status: Active, not recruiting

Conditions: UTUC

Treatments: Other: Deep learning system for prognostication prediction in upper tract urothelial carcinoma

Study type: Observational

Funder types: Other

Identifiers

NCT06993779

AI-UTUC

K2024-187-07 (Other Identifier)

Details and patient eligibility

About

Upper Tract Urothelial Carcinoma (UTUC), characterized by its anatomical complexity and often aggressive clinical behavior, presents substantial difficulties in accurate diagnosis and reliable prognostication. The stratification of postoperative survival utilizing radiomics features derived from imaging and characteristics from whole slide images could prove instrumental in guiding therapeutic decisions to enhance patient outcomes. In this research, our objective is to construct a deep learning-based prognostic-stratification system designed for the automated prediction of overall and cancer-specific survival in individuals diagnosed with UTUC.

Full description

Upper Tract Urothelial Carcinoma (UTUC) can be challenging to accurately diagnose and its course difficult to predict, as the disease manifestations and aggressiveness can differ significantly among individuals. This research seeks to create an innovative system employing artificial intelligence to process patient data, encompassing images from diagnostic scans and surgical pathology slides. This system would then be capable of automatically forecasting a patient's overall survival and their specific likelihood of surviving UTUC. Such insights could empower clinicians to tailor more effective treatment strategies for each individual patient.

Enrollment: 1,000 estimated patients

Sex: All

Volunteers: No Healthy Volunteers

Inclusion criteria

• Patients with Upper Tract Urothelial Carcinoma (UTUC) who had radical nephroureterectomy (RNU).
• Contrast-enhanced CT scan (e.g., CT urography) less than two weeks before surgery.
• Complete CT image data and clinical data.
• Complete whole slide image data.

Exclusion criteria

• Patients with a postoperative diagnosis of non-urothelial carcinoma.
• Poor quality of CT images and/or whole slide image data.
• Incomplete clinical and follow-up data.

Trial design

1,000 participants in 1 patient group

AI-UTUC

Description:

Patients with Upper Tract Urothelial Carcinoma (UTUC) who underwent radical nephroureterectomy (RNU)

Treatment:

Other: Deep learning system for prognostication prediction in upper tract urothelial carcinoma