This AI-driven prognostic and predictive model aims to predict outcomes for muscle-invasive bladder cancer (MIBC) patients undergoing neoadjuvant chemo-immunotherapy, aiding in treatment decision-making and patient stratification.
The market for bladder cancer diagnostics lacks efficient prognostic tools that can accurately predict treatment response and pathologic downstaging. Current risk classification systems exhibit limitations in identifying non-responders to neoadjuvant therapies, resulting in suboptimal treatment outcomes. In the US, bladder cancer is the most common cancer in men with 83,000 new cases diagnosed every year in men and women. 25% of bladder cancers are MIBC. There is an urgent need for more effective prognostic tools that employ modern technologies (AI/ML) systems to reliably discern these tumors quickly and accurately.
Utilizing machine learning algorithms, this artificial intelligence (AI) model analyzes histopathological images of muscle-invasive bladder cancer (MIBC) tumors to extract nuclear morphology and architectural features. These features are then used to predict treatment response and pathologic downstaging, enabling personalized treatment strategies for MIBC patients. The system delivers results from the feature-driven model allowing for a simple interface to be used by healthcare professionals. This method has been validated in a multi-institutional setting.
The AI-driven prognostic model has undergone validation in a single center Phase II trial (BLASST-1) and demonstrated promising predictive accuracy in identifying non-responders to neoadjuvant chemo-immunotherapy.