AI-Driven Integration of Pathologic, Radiomic, and Clinical Features for Prognostic and Treatment-Response Prediction in Gliomas

Dr. Zhao is a biomedical data scientist who hopes to use computational pathology and biostatistics to create a better way to manage and treat patients with central nervous system (CNS) gliomas or brain tumors. CNS gliomas are associated with high mortality and while long term outcomes are almost always poor, disease progression is highly variable, meaning that it can be hard to predict how the tumor will grow and at what rate.

With this Young Investor Award, Dr. Zhao will develop and validate a multimodal AI framework to support clinical decision-makings in adult diffuse gliomas. Using retroprosectively collected data, Dr. Zhao hopes to create an AI model that can predict how quickly a patient’s disease will progress or how long cancer may return after surgery. Additionally, Dr. Zhao suggests that an AI model that integrates histological, radiological and clinical features could produce individualized risk scores that would assist neuro-oncologists to predict post-operative treatment responses and, manage potential risk. Beyond predictive performance, Dr. Zhao hopes a working and proven model may be extended to serve as a blueprint for AI models to predict the progression of other brain tumors and solid tumors in general.