Our expert answers 3 Questions
My lab unites the expertise of computational and mathematical experts working with clinical researchers across all areas of medicine. Our goal is to match the many big and complex problems in clinical care with tools developed through computational methods that may offer solutions to these problems.
Our work is specifically focused on developing computer-assisted decision-making systems. These systems can assist providers and caregivers by offering real-time treatment recommendations and outcome predictions at every stage of care. These systems are developed using signal processing, image processing machine learning, and other techniques to extract important information from a myriad of clinical data sources, which is then communicated in the form of recommendations and predictions through easy-to-understand user interfaces to help providers make optimized decisions.
One of the big issues facing medicine is not a lack of data, but the opposite problem – healthcare generates a tremendous amount of data, which is extraordinarily complex on a level that is difficult for humans to comprehend and overwhelms decision-making capabilities. The computational algorithms within a clinical support system can synthesize all this diverse and heterogeneous data available about each patient at any given time, determine patterns, and continuously update and refine predictions based on the latest information available, including the aggregate expertise of clinicians.
As an example, one of our projects is developing a system to monitor people recovering from cardiac surgery in the hospital in real time to try to predict potential complications that may be on the horizon. We’re also using similar algorithms to monitor cardiovascular health after patients go home, utilizing data generated through wearable devices and other sources.
Computer-assisted decision-making systems have the potential to improve care, reduce costs, and provide expertise in situations or settings where it might not immediately be available.
While these systems have been an integral part of healthcare for many years, recent advances in computational methods that can harness the richness within the broad types of patient data available today make this a critical moment for these tools to make significant contributions toward improving clinical care and patient outcomes.
Medicine will always be a fundamentally human experience dependent on relationships between patients and providers. As useful as these tools can be, at the end of the day they serve in an advisory function to the teams of people who ultimately make the decisions. A completely automated decision support system will not happen in foreseeable future, but these tools will significantly change providers’ roles in many fields.