This lab is committed to developing advanced computational methods to address complex problems in medicine and biology. Research projects are driven by a wide range of federal and private organizations, including the American Heart Association, the National Science Foundation, the Department of Defense, the Michigan Center for Integrative Research in Critical Care (MCIRCC) , and Michigan Medicine.
The philosophy behind our efforts is threefold: enjoy your work, satisfy your scientific curiosity by developing and applying innovative and sophisticated methodologies, and help save lives by creating new and novel solutions!
Computational Medicine: A Gateway to Computer-Aided Decision-Making
The focus of the research conducted in our lab is in the development of computer-assisted decision-making systems. Using insightful bio-markers and underlying patterns extracted from raw data via signal processing, image processing and machine learning techniques, these systems can provide real-time, on-the-fly treatment recommendations and outcome predictions at every stage of care. By assisting physicians in this manner, patient care can be improved, costs can be reduced and expertise can be provided in cases where medical experts are not immediately available. Example applications of these systems are in cases of traumatic injury and hemorrhagic shock; where all available structured and un-structured health data of the patient is processed as and when it’s made available to produce diagnostic, predictive and prescriptive analysis of the patients conditions in real time. The raw data we analyze includes time-series waveform data such as ECG, ABP etc., medical imaging data such as CT, X-Rray, MRI & fMRI, etc., and all other related patient centric information such demographic information, EHR variables, lab results etc.
These recommendations and predictions are then communicated to physicians and caregivers via simple, effective, purpose designed user-interfaces in an easy-to-understand and interactive format, along with clear and descriptive visualizations, which communicate the computed outputs and the reasoning behind it reasoning.
Computational medicine encompasses the techniques that apply mathematical tools to extract important diagnostic information from biomedical and biological data. Data mining techniques can then be applied to train expert systems on existing data and classify new examples. The core research and development of the lab is primarily focused on developing novel mathematical techniques of signal processing, image processing and machine learning towards developing novel solutions for medicine. However, due to the size, complexity and the disparate nature of medical data, we also use and develop variety to techniques for data wrangling, data imputation, data aggregation, alignment, multi-core multi-threaded processing, distributed computing, etc. which help us develop effective, robust and commercial grade medical discussion assist systems.