Douglas E. Lake, PhD

Douglas E. Lake, PhD, is a mathematician/statistician and data scientist. He received his BS in Applied Mathematics from the University of Virginia (UVA) School of Engineering in 1982 and a Ph.D. in Mathematics from UVA in 1991. 

Since joining the UVA faculty in 1999, Lake has worked closely with Randall Moorman in rigorously developing, optimizing, and applying mathematical tools for clinical prediction problems.  He has played a significant role in developing the HeRO system and evaluating this and other heart rate characteristics (HRC) algorithms for the early detection of sepsis in premature infants. Lake has provided statistical and signal processing expertise to numerous other NIH-sponsored projects; his primary role is developing robust mathematical and scientific approaches to analyzing bedside monitoring data. His role includes managing and processing vast amounts of time series data across intensive care units at UVA and multiple other hospitals.

He is currently the lead data scientist for the UVA Neonatal ICU Research Group, focusing on developing and validating POWS (pulse oximetry warning score) to predict sepsis at UVA and several other sites. He is co-PI in UVA’s role as Leadership and Data Coordination Center for the NIH-funded U01 project Pre-Vent. This recently completed study measured the maturation of ventilatory control and tested its association with clinical respiratory outcomes in over 700 extremely premature infants at 5 NICUs. This included developing and implementing signal-processing algorithms for studying apnea, hypoxemia, and periodic breathing.

Lake developed the coefficient of sample entropy, which included using entropy rate measures to detect atrial fibrillation in very short heart rate records.  He has provided other new theoretical and statistical analyses of related entropy measures for time series.  More recently, he has developed highly comparative time series analysis techniques for the discovery of new signatures of illness in cardiorespiratory physiology for predicting mortality, sepsis, respiratory outcomes, cerebral palsy, and autism.