Unlocking the Potential of Big Data to Personalize Patient Care
New ROI Grant Fills a Funding Gap for Emerging Field
How would practice change if every radiation oncology patient could be treated as if they are in a clinical trial and it would just happen as part of the standard work flow? Imagine the possibilities of what could be learned by analyzing all of that data! The Radiation Oncology Institute’s (ROI) newest grantee, Todd McNutt, PhD, is a lead researcher on Oncospace, a complex system that assembles and analyzes data from the treatment planning system and patients’ clinical records and can do just that. With his recent $200,000 award from ROI, Dr. McNutt is using data from Oncospace to investigate how big data analytics and machine learning approaches can be used to develop better predictive models and decision support tools for more personalized treatment planning and clinical interventions for head and neck cancer patients.
Dr. McNutt and his team at The Johns Hopkins University have been developing Oncospace since 2007. Their initial focus was on studying how radiation oncology patients moved through their system and building tools that would facilitate collecting patient data at the point of service and in a useable format. Now, they have established a foundation for a learning health system that improves with each new patient entered, added three more institutions to form an Oncospace consortium and amassed a large amount of clinical data.
Oncospace is forging a pathway to analyzing the vast amount of knowledge and experience contained in prior clinical records that has otherwise been inaccessible. “What we can do using big data and machine learning algorithms could be far more powerful and impact clinical care faster than traditional clinical trials,” says Dr. McNutt. “The key is predictive modeling that starts a cycle. We want to use the computer to recall and process the experiences and outcomes from a large pool of prior patients to provide physicians with decision support tools as they care for new patients. Over time, each new patient added will help improve the predictions and decision support provided and result in better outcomes for patients.”
For the research supported by ROI, Dr. McNutt will collaborate with radiation oncologist Harry Quon, MD, who specializes in treating head and neck cancers and has provided a critical link to the clinical setting throughout the development of Oncospace. The team will refine and expand their predictive models of treatment toxicities related to weight loss for head and neck cancer patients and incorporate spatially dependent features of the radiation dose distribution to better understand how the spatial regions of organs at risk may be related to the toxicities. They also aim to produce a clinically viable model of a decision support system that would provide physicians with a prediction of toxicity for a new patient along with personalized, actionable information. Eventually, they would like to make such tools accessible to any physician through an Oncospace web portal.
Despite its promise to improve the quality and safety of radiation oncology, Oncospace doesn’t fit the mold of projects that are typically supported by more traditional research funding sources. The ROI selected Dr. McNutt’s proposal for funding from 16 applications that were submitted in response to its “Leveraging Big Data to Optimize Quality Assurance and Patient Care Improvement Initiatives” request for proposals that sought to encourage research in this emerging field. “The ROI award is an important recognition by our peers of the work that we are trying to do,” says Dr. McNutt.
For more information, visit the Oncopace website.