AI Diagnostic Tool Improves Cancer Detection in Cystoscope Images of Bladder

It is known to have high recurrence rates and significant risks of disease progression. Early detection of bladder cancers and recognition of disease recurrence can substantially reduce patient morbidity and healthcare costs, reduce the risks of disease progression, and improve overall survival. Now, a new image enhancement and artificial intelligence (AI) diagnostic tool for bladder cancer detection in images (videos and camera stream) seen on white light cystoscopy and narrow band imaging (NBI) cystoscopy could be beneficial in improving bladder cancer diagnostics and patient care.

Claritas HealthTech (London, UK) is commencing clinical validation of CystoSmart, a Software as a Medical Device (SaMD), that has been jointly developed through a research collaboration with Khoo Teck Puat Hospital (KTPH, Singapore).

“Our aim has been to develop an AI diagnostic adjunct that enhances the accuracy of detection of bladder cancers. This will be beneficial in allowing: 1) Appropriate treatment for newly diagnosed cancers 2) Accurate recognition of tumor recurrences 3) Complete tumor resection during surgery,” explained Dr. Yeow Siying, a consultant urologist at the Department of Urology, KTPH, who is the medical principal investigator for the project. “The detection of bladder cancer often involves cystoscopy, where a fibre optic camera is inserted into the bladder to visualize its inner lining (mucosa). Most commonly, white light cystoscopy is utilized, whilst other adjuncts such as Narrow Band Imaging (NBI), can help improve the accuracy of cancer detection to a limited extent. Detection of bladder cancer can be challenging, particularly for flat lesions such as carcinoma-in-situ. Certain benign conditions may appear visually similar to bladder cancers as well.”

“Training AI tools that are reliable in clinical settings have been challenging because it is dependent on the quality of the captured cystoscopic images. Claritas has leveraged its image enhancement platform to create enhanced, clinically validated data sets, thereby allowing more detail to be extracted in training our neural network,” said Dr. Arup Paul, Chief Clinical Strategy Officer at Claritas. “Our pre-clinical studies have allowed the appropriate calibration and development with resulting evidence of high sensitivity and specificity. We are confident of moving into the next stage, with clinical evaluations utilizing CystoSmart to demonstrate benefits for patients, clinicians, and healthcare systems.”

Related Links:
Claritas HealthTech 
KTPH