Director: Yevgeniy Vorobeychik, Ph.D.
Our research focuses on blending economic and computational methods to tackle a broad array of socially relevant problems, including a data-driven investigation of innovation diffusion (with a focus on diffusion of sustainable energy technologies), game theoretic modeling and analysis of privacy and security risk, and game theoretic modeling and bi-level optimization methods for antibody design.
Computational Medicine Laboratory
Director: Thomas Lasko, M.D., Ph.D.
Our research is focused on improving clinical diagnosis, prognosis, and treatment decisions by applying mathematics and computational science to the large datasets collected in routine clinical practice. For example, we are working to identify and extract signals from population-scale Electronic Medical Records data that can provide clinically useful visual feedback about individual patient disease states to practicing anesthesiologists. We are also working hard on the phenotype discovery problem, one of the key computational problems underlying precision medicine - it is the effort to computationally infer all existing phenotypes (including previously unrecognized diseases and disease subtypes) from population-scale clinical data. Our general approach is to use unsupervised, data-driven machine learning methods to discover the important signals and patterns in data, remaining as free as possible from human bias and cognitive blind spots.
Director: Tony Capra, Ph.D.
Humans differ from one another and our closest living relatives, the chimpanzees, in a wide range of traits, including our susceptibility to many diseases. We model the evolutionary processes that have produced these novel traits and develop algorithms that compare genomes to predict the functional relevance of specific genetic differences between individuals and species. Our research is motivated by several questions:
1) How have evolutionary processes produced the astonishing diversity of form and function present in the natural world?
2) How can better algorithms lead to a deeper understanding of biological systems and networks?
3) How do genomes encode and maintain the information necessary to produce life?
4) How can our increasing knowledge of genomic variation be translated into the treatment and prevention of disease?
We investigate these questions in a number of model systems, but our main focus is on the origins and recent evolution of human populations and their primate relatives.
The GetPreCiSe Center is an NIH Center of Excellence in Ethics Resarch. We founded the center based on the observation that the debate about genetic privacy and identity has been based (a) on an incomplete understanding of the influences on the actors involved in genomics research and translation and (b) on possible, rather than probable, risks. Moreover, research has typically focused on what individuals say, effectively minimizing the role of community and social influences in shaping attitudes toward privacy. GetPreCiSe integrates a diverse group of interdisciplinary scholars and community advisors to collaborate and develop a more comprehensive understanding of these worries and the factors that influence them, to model actual risks to privacy and identity, all of which will be used to inform policy.
Director: Daniel Fabbri, Ph.D.
We founded the HAILLab to address the challenges at the intersection of: (i) computer science, machine learning and data management, and (ii) healthcare. We develop and apply computational methods to a wide range of healthcare issues from discharge prediction to cancer recovery times. Additionally, we develop hospital IT infrastructure such as patient engagement systems and medical record search engines to improve health data management. Our lab members work to publish papers in computer science and biomedical informatics proceedings, and deploy tools in hospitals and the health community.
Director: Colin Walsh, M.D., M.A.
The HARBOR Lab is a multidisciplinary team of investigators based in the Department of Biomedical Informatics with expertise in clinical medicine, healthcare quality, engineering, and computer science. Our research focuses on the application of data science to large scale clinical data to enable population health. Specifically, their research includes 1) Clinical Predictive Modeling, 2) Interactive Data Visualization, and 3) Natural Language Processing, In particular, we focus on applications in mental and behavioral health, with an emphasis on modeling and risk stratification.
Director: Bradley Malin, Ph.D.
The HIPLab was founded to address the growing needs for privacy technology research and development for the emerging health information technologies sector. Our goal is to improve the protection of patients' privacy in health information systems. Our research is rooted in basic science and and application in a number of health-related areas, including primary care and secondary sharing of patient-specific data for research purposes. Projects in the HIPLAB are multi-faceted and draw upon methodologies in computer science, the biomedical sciences, and public policy. We aim to improve the standard of healthcare and health information systems by developing technologies that enable trust.
Director: Bennett Landman, Ph.D.
Three-dimensional medical images are changing the way we understand our minds, describe our bodies, and care for ourselves. In the MASI lab, we believe that only a small fraction of this potential has been tapped. We are applying medical image processing to capture the richness of human variation at the population level to learn about complex factors impacting individuals. Our focus is on innovations in robust content analysis, modern statistical methods, and imaging informatics. We partner broadly with clinical and basic science researchers to recognize and resolve technical, practical, and theoretical challenges to translating medical image computing techniques for the benefit of patient care.
Director: You Chen, Ph.D.
Our lab focuses on the development of methods to learn workflows, health processes, and models the organizational structure (e.g., via social network analysis) of complex health systems. We also apply these methods to investigate how best to optimize the practices and resource allocation to improve care quality for patients and minimizes healthcare costs.
Statistical Evidence in Data Science Laboratory
Director: Jeffrey Blume, Ph.D.
Our research focuses on three areas: (1) Likelihood methods for measuring statistical evidence / foundations of statistical inference, (2) methodology for analyzing and interpreting Receiver Operating Characteristic (ROC) curves, and (3) the design, conduct and analysis of Clinical Trials. We work to bring likelihood methods to bear on statistical aspects of the drug/device approval process and to develop likelihood methods that will yield highly efficient adaptive clinical trial designs.