Awarded $250,000 from the Cancer Digital Intelligence’s Grand Challenge, Hopkins aims to improve treatment options for brain and spine cancers
By Elizabeth Kleisath
Communications Officer, Advancement
This year, approximately 3,300 Canadians will be diagnosed with cancer in their brain or spine. Of these patients, 79 % will die from their central nervous system (CNS) cancers, due to the limitations of our current treatment options. Dr. Scott Hopkins, a professor in the Department of Chemistry at the University of Waterloo, is working to improve these survival rates.
CNS cancers include many different types of tumors. In current treatment methods, patients undergo neurosurgery to remove the tumor, which is then sent for post-operative analysis to determine the tumor type. However, different types of CNS cancers have individual surgical guidelines based on how aggressive they are. This means that during this initial surgery, doctors do not know which guidelines to follow to balance removing the tumor and preserving all the healthy tissue, and patients must then undergo multiple neurological surgeries to meet the treatment guidelines.
Hopkins, in partnership with collaborators at Princess Margaret Hospital, is developing a new diagnostic tool that can identify the CNS cancer type in just 10 seconds. With this short time frame, it could be used mid-surgery to provide surgeons with accurate information. Their project has just been selected as the winner of the Cancer Digital Intelligence (CDI) Grand Challenge: Responsible Artificial Intelligence, and the team has been awarded $250,000 in resources to implement this project.
“This is a really exciting new technology that will provide surgeons and pathologists with real-time diagnostic feedback during surgeries,” says Hopkins. “Our hope is that one day the AI-driven PIRL-MS technology will improve cancer patient outcomes across Canada.”
Hopkins will collaborate with Dr. Arash Zarrine-Afsar, co-inventor of the picosecond infrared laser mass spectrometry (PIRL-MS) diagnostic tool, to develop an AI that identifies biomarkers specific to a variety of cancers. For example, PIRL-MS can obtain a “fingerprint” of fatty compounds known as lipids in the tumor mass, which are altered in cancer patients and are thus biomarkers of the disease. These fingerprints are then compared to known tumor samples for identification. This process itself is not new—researchers can currently differentiate cancers with 95% accuracy—however, for use in clinical settings, the accuracy must be improved to over 98%.
Hopkins is improving the accuracy by integrating machine learning into PIRL-MS. By developing a new classification model for CNS cancers, the artificial intelligence can help manage the complex datasets, provide insights into rare cancer types, and use decision tree relationships to diagnose the cancer type and provide accurate treatment recommendations.
Dr. Arash Zarrine-Afsar, University Health Network
The team’s proof-of-principle has already provided diagnosis for 156 brain cancer specimens, and the team is curating their extensive library of over 21,000 cancer “fingerprints” for integration into new machine learning algorithms.
“We are grateful for the support from Cancer Digital Intelligence,” says Hopkins. “Leveraging the power of AI with the fast and accurate molecular fingerprinting afforded by PIRL-MS will transform our ability to treat brain cancer.”
This innovation in biomedical technology seeks to meaningfully improve patient outcomes, reduce healthcare costs, and increase healthcare equity by enabling more accurate diagnosis which can lead to single-surgery care and tailored treatments.
For those inspired to contribute to this crucial work or other cutting-edge advances in health research currently being pursued in the Faculty of Science, we invite you to connect with the Director of Science Advancement to learn how your generosity can make a meaningful impact. Philanthropic support is transformative in advancing scientific research and enabling groundbreaking discoveries that can change lives.