“Knowledge Discovery for Critical Care: a MIMIC-II Story”
by Mengling 'Mornin' Feng, Thomas Brennan and Leo Anthony Celi
Monday, September 23, 2013
10:30 – 11:30 AM
Conference Room B (lower level)
Natcher Building (Bldg 45)
Abstract: The data generated in the process of medical care has historically not just been under-utilized, it has been wasted. This was due in part to the difficulty of accessing, organizing and utilizing data entered on paper charts, but notable variability in clinical documentation methods and quality made the problem even more challenging. In the absence of a practical way to systematically capture, analyze and integrate the information contained in the massive amount of data generated during patient care, medicine has remained a largely ad hoc process in which the disconnected application of individual experiences and subjective preferences continues to thwart continuous improvement and consistent delivery of best practices to all patients. The ICU presents an especially compelling case for clinical data analysis. The value of many treatments and interventions in the ICU is unproven, and high-quality data supporting or discouraging specific practices are embarrassingly sparse. Over the past decade, the Laboratory of Computational Physiology at the Massachusetts Institute of Technology, Beth Israel Deaconess Medical Center (BIDMC) and Philips Healthcare, with support from the National Institute of Biomedical Imaging and Bioinformatics, have partnered to build and maintain the Multi-parameter Intelligent Monitoring in Intensive Care (MIMIC) database. This public-access database, which now holds clinical data from over 40,000 stays in BIDMC ICUs, has been meticulously de-identified and is freely shared online with the research community via PhysioNet. Our vision is for the development of a care system consisting of “clinical informatics without walls”, in which the creation of evidence and clinical decision support tools is initiated, updated, honed, and enhanced by crowd sourcing. In this collaborative medical culture, knowledge generation would become routine and fully integrated into the clinical workflow. This system would employ individual data to benefit the care of populations and population data to benefit the care of individuals.
Mengling 'Mornin' Feng, Ph.D.: Dr. Mengling Feng (http://web.mit.edu/mfeng/www/) is currently a Post-Doc Fellow in the Lab of Computational Physiology, Harvard-MIT Health Science Technology Division. Dr. Feng obtained both his Bachelor and PhD degrees from School of Electronic and Electrical Engineering, Nanyang Technological University. Under the supervision of Prof. Limsoon Wong (School of Computing, NUS) and Prof. Yap-Peng Tan (School of EEE, NTU), Dr. Feng’s Ph.D. study focused on developing data mining methods to discover meaningful knowledge that impacts real life practices. Before his current affliction at MIT, Dr. Feng joined the Data Analytic Department of Institute for Infocomm Research (I2R) as a research scientist. Dr. Feng was awarded the Ministry of Education Scholarship for his undergraduate studies and the A*STAR Graduate Scholarship for his PhD study. His work was also recognized with the “Bi-annual Best Paper Award” from the Institute for Infocomm Research. Dr. Feng’s research focus is to develop data mining and machine learning methods to discover or infer casual phenomenon among real-life practice and strategic planning.
Thomas Brennan, Ph.D.: After completing a Bachelors in electrical and computer engineering at the University of Cape Town, Thomas Brennan was awarded the Rhodes Scholarship from South Africa in 2004. In 2009 he completed his D.Phil. in Biomedical Engineering from the University of Oxford. He then worked for the Vodafone Foundation developing a mobile-based platform to monitor and support community health workers in South Africa. In 2010, he accepted a Wellcome Trust post-doctoral research fellowship at the Institute of Biomedical Engineering in Oxford to develop and assess mobile health solutions for monitoring chronic disease in resource-constrained settings. In 2012, he took up a Research Engineer at the Laboratory of Computation Physiology at Massachusetts Institute of Technology, where he is now working on developing systems to monitor patients in intensive care unit. He has over 10 years experience in clinical informatics, biomedical signal processing, with a special focus on machine learning.
Leo Conti, M.D.: Leo moved to the US from the Philippines after medical school to pursue specialty training in internal medicine (Cleveland Clinic), infectious diseases (Harvard) and critical care medicine (Stanford). He has practiced medicine in three continents (Philippines, US and New Zealand) and has worked in both industry (Philips Visicu) and academe (faculty positions at Harvard, MIT, Stanford and University of Otago), rendering him with broad perspectives in healthcare delivery. He has a strong interest in systems re-design for quality improvement, and became the New Zealand representative to the Quality and Safety Committee of the Australia New Zealand Intensive Care Society in 2006. Feeling he needed more skills to tackle the healthcare inefficiencies he faced wherever he practiced, he went back to the US to pursue graduate studies in biomedical informatics at MIT and public health at Harvard. While attending both schools and working part-time as an emergency department physician, he co-founded Sana, personally recruiting most of the current members, and was instrumental in shaping the mission and vision of the young organization. His other research interest is in data mining and the application of machine learning on large databases. As a research scientist at the Laboratory of Computational Physiology at MIT, he works with MIMIC, a publicly-available de-identified ICU database from BIDMC. He is working on a data-driven decision support system known as Collective Experience that (1) allows a clinician to draw on the experience of other clinicians who have taken care of similar patients as recorded in a clinical database, and (2) uses models performed on relatively homogeneous patient subsets.
For more information, please contact Jim DeLeo, firstname.lastname@example.org, 301-496-3848.