AI-guided ultrasound and MRI staging to guide bladder cancer treatment

Intelligent Diagnosis of Bladder Cancer Staging and Prediction of New Adjuvant Chemotherapy Efficacy Based on Deep Learning and Transfer Learning in Ultrasound-Magnetic Resonance-Pathology Multimodal Multiscale

This project will test whether a deep-learning system using ultrasound, MRI, and pathology can more accurately stage bladder cancer and predict who will benefit from additional chemotherapy for patients with suspected bladder tumors.

Quick facts

• Study type: Observational
• Enrollment: 480 (estimated)
• Sex: All
• Sponsor: Sun Yat-Sen Memorial Hospital of Sun Yat-Sen University (Academic / other)
• Drugs / interventions: chemotherapy, radiation
• Locations: 1 site (Guangzhou, Guangdong)
• Trial ID: NCT07051083 on ClinicalTrials.gov

What this trial studies

Researchers will build a multimodal dataset linking contrast-enhanced ultrasound, magnetic resonance images, pathology slides, clinical data, and follow-up for patients with suspected bladder cancer. They will preprocess and homogenize imaging, perform tissue segmentation with U-Net and Transformer-based methods, and train convolutional neural networks with attention mechanisms and transfer learning to derive staging and prognostic models. Tumor specimens will undergo immunohistochemical fluorescence staining to add molecular information and support multimodal feature extraction. The project is observational and will enroll surgical candidates at a single center to test the models' ability to stage disease and stratify risk.

Who should consider this trial

Good fit: Adults with suspected primary or recurrent bladder cancer who are planned for surgical resection, have not received prior radiotherapy or chemotherapy, and can undergo contrast-enhanced ultrasound and MRI are ideal candidates.

Not a fit: Patients who cannot tolerate or are allergic to ultrasound contrast agents, cannot complete the required imaging, have other primary malignancies, or have already received prior treatment may not benefit from this project.

Why it matters

Potential benefit: If successful, the system could provide more accurate preoperative staging to reduce overtreatment and help clinicians choose more appropriate surgery or chemotherapy.

How similar studies have performed: Previous AI and imaging studies have shown promising results for bladder cancer staging, but the specific multimodal combination of ultrasound, MRI, pathology, and transfer-learning deep models remains largely novel and unproven.

Eligibility criteria

Inclusion Criteria:

1. Ultrasound and other imaging examinations (CT, MR, etc.) suggest bladder masses and are suspicious for bladder cancer patients.
2. The bladder is well filled, and no allergic reactions to ultrasound contrast agents are found.
3. No surgery or radiotherapy/chemotherapy has been performed.
4. Patients who meet the indications for surgical resection and are planned for surgical treatment, including one of the following:

   1. Clinical symptoms consistent with suspected bladder cancer (such as gross hematuria, etc.);
   2. Patients with confirmed primary or recurrent bladder cancer by cystoscopic biopsy;
   3. Rapid urine cytology and urine cytology FISH testing suggest malignancy.

Exclusion Criteria:

1. Individuals unable to tolerate surgery;
2. Individuals allergic to ultrasound contrast agents, unable to undergo ultrasound contrast examination;
3. Unsuccessful preoperative ultrasound contrast examination or non-compliant patients;
4. Postoperative pathology does not indicate bladder cancer;
5. Patients who have undergone chemotherapy or radiation therapy.

Where this trial is running

Guangzhou, Guangdong

• Sun Yat-sen Memorial Hospital, Sun Yat-sen University — Guangzhou, Guangdong, China (Recruiting)

Study contacts

• Principal investigator: Qiyun Ou, Dr. — Sun Yat-Sen Memorial Hospital of Sun Yat-Sen University
• Study coordinator: Qiyun Ou, Dr.

Conditions Bladder Cancer; Staging; Deep Learning; Neoadjuvant Chemotherapy; Contrast Enhanced Ultrasound; Multi-omics