Research - UNCC Health Informatics Lab

Research Scope:

Biology and medicine are still in the rudimentary stage of accumulating observations and knowledge piece by piece. Scientists desire to discover new universal laws and principles like Darwin's theory, based on tremendous observational data scattered all around. I strongly believe that data-driven computational approaches, complementary to classical hypothesis-driven approaches, have the potential to go beyond any single human expert's capability of comprehension and interpretation. They are very useful for discovering patterns or rules and filling ignored niches in the current biomedical knowledge system.

Our long-term research goal is to develop and apply data mining technologies to uncover new clinical knowledge and to improve the quality, safety, efficiency and effectiveness of healthcare.

Current Research Projects:

Behavioral Health Stigma - Breaking the Code with Stigma Index (Dr. Hadzikadic)
  • Social labeling of individuals comes in many forms under the umbrella known as “Stigma”. Stigma can be the leading cause for limiting the access of many individuals to care, which can put constraints on their ability to integrate easily in the society. Behavioral health stigma is one of the different forms that stigma can take. Stigma Index (SI) will serve as a tool to measure and quantify the impact of stigma over time and across different societies as perceived by: individuals with behavioral health illness, the communities to which they belong to, and psychiatrists in behavioral health clinics. The SI comes along the lines of the Consumer Sentiment Index (CSI), which is a statistical measurement of the economy’s overall health as determined by the consumer opinion. While the CSI reflects the overall perception about economy, the SI reflects the overall perception about behavioral health stigma across the United States.
Cardiovascular Research Grid - Imaging Informatics (Dr. Ge)
  • The CardioVascular Research Grid (CVRG) project is creating an infrastructure for secure seamless access to study data and analysis tools. CVRG tools are developed using the Software as a Service model, allowing users to access tools through their browser, thus eliminating the need to install and maintain complex software. The CVRG project is supported by the National Heart Lung & Blood Institute. The project is based at the Institute for Computational Medicine at the Johns Hopkins University, Department of Biomedical Informatics at Vanderbilt University Medical Center, The College of Computing and Informatics at UNC Charlotte, The Center for Comprehensive Informatics at Emory University, The College of Engineering and Applied Sciences at Stony Brook University, and the Computation Institute at The University of Chicago (CVRG Research & Development Team).
    The Imaging Informatics program within CVRG focuses on developing platforms, data, and resources to support longitudinal population studies for cardiovascular research that rely extensive on imaging-based phenotypes.
  • This project is funded by a grant from NIH/NHLBI.
Dataset Information Resources (DIR) - An Information Framework for Discovering and Learning Datasets (Dr. Ge, Dr. Yao)
  • In this project, we aim to develop a web-based and eventually mobile-driven information framework for discovering and learning datasets that helps students and researchers in data science. A prototype of the framework has already been developed, which is accessible via
    As recognizing the importance of available data for effective research in the era of big data, many data portals have been developed to provide data of various types for query, analysis, and download by citizens and researchers alike. Most of the existing data portals provide some instructions and tools for identifying and using datasets that are within individual portals. However, there is a lack of information resources that catalog and curate these datasets in a manner that facilitate research questions. This is especially true for new students and researchers entering this field. For them it is frequently a considerable challenge to find out what datasets are available, what data elements may be relevant to their research questions, and where and how they can obtain them. Particularly in health care, there are tens of thousands of publically available datasets and numerous proprietary datasets that can be purchased that are critical for health data analytics, but many of these datasets are in various difference places and often in very different formats. Thus, it is an important and urgent need to fill the gap and develop an effective information framework for disseminating knowledge about available datasets in data science.
A decision support model for pharmaceutical research prioritization (Dr. Yao)
  • Governmental and private funding agencies, pharmaceutical and biotech companies, different patient groups, and hundreds of thousands of scientists across the world constantly face the question – “Given limited time and resources, which disease area(s), or what unmet medical need(s) shall I invest in?” This is a hard decision making problem because it involves multiple factors in complex and dynamic ways. Currently ad hoc studies based on limited data are carried out in individual disease areas as common practice. There are no quantitative and comparative studies across the whole disease landscape to accurately address the problem. In this project, we propose to develop a decision support model for pharmaceutical research prioritization. We will apply it onto our data collected from a large medical claim database, a scientific literature database and the FDA clinical trial registry during 2000 and 2011. The goal is to systematically survey the resource allocation situation for about 1,500 diseases in the United States during that period, and to identify the ignored niches for future research by investigating the inequalities among disease burden, basic biomedical research, clinical development and current clinical practice.
Decision support for dose prescription in Intensity Modulated Radiation Therapy (Dr. Ge)
  • Intensity Modulated Radiation Therapy (IMRT) is an important clinical procedure for treating various cancer. It's estimated that more than half of all cancer patients receive radiation therapy as part of their treatment independently or in combination with chemo and other treatments. Due to its ability to conform radiation dose to the tumor while sparing surrounding normal tissue, IMRT has become widely used to treat ever more types of cancers in recent years. Unfortunately, the complexity of IMRT makes it costly requiring significant amount of planning efforts from clinicians and planners and yet it is still difficult to ensure high quality. This project aims to develop decision support tools to enable clinicians and planners to dramatically improve the efficiency and quality of IMRT planning.
  • This project is funded by a grant from NIH/NCI.