Grant: Alliance Grants

Integrating cutting-edge AI with human-centered clinical expertise plays critical roles in guiding clinical decisions in precision medicine, where doctors can make accurate diagnosis with that information to develop an individualized treatment plan for cancer patients. Advanced cancer genomics profiling has identified many clinically and biologically significant signatures that cannot be detected through traditional morphological examination or under a microscope for cancer diagnosis. AI and advanced big data analytics would be instrumental in finding clinically relevant signatures. Our goal is to develop AI-powered Intelligent Agents (IA) with smart chatbot to enhance clinical decision-making for glioma, an adult brain tumor, within a Tumor Board setting. The Tumor Board, a group of specialists who discuss complex cancer cases, will use this AI system as a supportive tool to guide clinicians in decision making process, specifically help clinicians make informed decisions on glioma diagnosis, prognosis, and treatment. Large Language Models (LLM) could understand human-made texts (prompts) that makes it easier to interact with clinicians. We will develop LLM-powered IAs to automate data collection, processing, and reporting. The specific objectives are: 1) to create the core components of the IA; 2) to integrate the IA with LLMs and develop a chatbot interface for clinical support; and 3) to deploy the IA with LLM and assess its performance. The established collaboration with an AI expert from industry, a pathologist from hospital Tumor Board and Software/Biomedical Engineer from academia will facilitate the translation of the novel AI strategy into the clinical practice. Overall, our proposal can make a significant contribution to streamline diagnostic decision-making processes and improve glioma diagnosis by resolving diagnostically ambiguous difficult cases, and the efficiency of overloaded clinicians in a busy clinic which will lead to huge healthcare benefits to Canada. Furthermore, this AI system can be adapted to other cancer types or diseases that potentially transform clinical diagnostics. It will also help Canada to establish a strong leading position in AI-driven healthcare solutions.