David R. Holmes III, Ph.D., studies the methods by which health care researchers analyze and interpret large data sets. Health-related data is constantly being collected through the digitalization of medical records and the concordant interest in personal health monitoring. Because of this wealth of information, health care data makes up significant sectors of the big data industry, including in-hospital sensing, clinical reporting, medical imaging and home-based wellness.
Health care data is large, diverse and often incomplete, and is difficult to analyze and interpret. Dr. Holmes is investigating new methods, which attempt to link disparate data sets and build computational models directly from the data.
Research into data representation, signal processing, graph analytics and machine learning is yielding new and novel insight into basic biology and human health. To match the complexity of these algorithms and the magnitude of data, Dr. Holmes also studies the mapping of health care questions onto novel computational architecture.
- Health data formation and representation
- Machine learning techniques
- Analysis of clinical and outcome data
- Image and data fusion
- Normative and disease-specific computational models
Significance to patient care
Dr. Holmes conducts his research from the perspective that data drives decisions. In the case of personal health, individual and population-based data can inform health decisions, disease characterization and therapy planning. More specifically, by mining the clinical and imaging data that exist within the electronic medical record, Dr. Holmes seeks to develop better-individualized strategies for patient health.
As these techniques mature, it is likely that physicians and patients will be increasingly guided by personalized health strategies, which yield the highest value to the patient.