Bladder cancer (BLCA), with its diverse histopathological features and varying patient outcomes, poses significant challenges in diagnosis and prognosis. Postoperative survival stratification based on radiomics feature and whole slide image feature may be useful for treatment decisions to improve prognosis. In this research, we aim to develop a deep learning-based prognostic-stratification system for automatic prediction of overall and cancer-specific survival in patients with BLCA.

Bladder cancer can be difficult to diagnose and predict outcomes for, as the disease can vary greatly between patients. This research aims to develop a new system that uses artificial intelligence to analyze patient information, including images from surgery and scans. This system could then automatically predict a patient's overall survival and how likely they are to survive specifically from bladder cancer. This information could be used by doctors to make better treatment decisions for each patient.

Inclusion Criteria:

• patients with bladder cancer who had surgery like radical cystectomy or transurethral resection of bladder tumour (TURBT)
• contrast-CT scan 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
• incomplete clinical and follow-up data

We included patients who had surgery only or who had neoadjuvant chemotherapy before surgery. We excluded patients with a postoperative diagnosis of non-urothelial carcinoma.

The datasets analyzed during the current study are not publicly available due to the privacy of patients but are available from the corresponding author on reasonable request.