Developing scalable mHealth big-data solutions
Graduate Student → Professor → Chief Software Architect
Sense - Analyze - Act
I build open source software and technology to support reliable high-frequency data collection from mobile and wearable sensors to enable sensor-triggered just-in-time adaptive interventions. This is coupled with an analytics cloud designed to facilitate mobile health data analytics and interventions that enables population-scale model development, training, and evaluation. It brings state-of-the-art research techniques and methodologies together in a single system to affect clinical outcomes.
I am the Chief Software Architect for the NIH sponsored Mobile Sensor to Knowledge (MD2K) center where I design and build high-performance software to process real-time mobile sensor data for medical and workplace applications.
I was an assistant professor of Computer Science at the University of Memphis where my research focused on mobile health (mHealth) applications, particulary wearable technology to address obesity and mobile medication adherence for asthma.
I am an amateur photographer who enjoys taking photos of my family, ballroom dancing, and random things.
I have participated in social and competitive ballroom dancing since 2002.
These studies evaluate the feasibility of a just-in-time intervention to delay or prevent smoking relapse in smokers attempting to quit and the efficacy of the novel EasySense wireless, contactless system in assessing pulmonary congestion via measurements of thoracic impedance and cardiac and lung motion in patients with congestive heart failure (CHF) during hospitalization and post-discharge.
This study utilizes wearable sensors to objectively assess everyday job performance for employers.
These studies the influence of socioeconomic status, social history, contextual and environmental influences, biobehavioral/psychosocial predispositions, and acute momentary precipitants on stress, smoking lapse, and abstinence among 600 smokers attempting to quit.
This study examines the multiple neurocognitive processes that have previously been implicated in relapse among smokers who are both successful and unsuccessful in maintaining smoking abstinence.
This feasibility study examines the effects of delivering mindfulness strategies via smartphones on key mechanisms underlying smoking cessation among low socioeconomic status, racially/ethnically diverse smokers.
This study develops the Remote Oral Behaviors Assessment System (ROBAS) by integrating a multimodal sensing platform (smart toothbrush and wrist sensors) with the mCerebrum software platform (physiological and EMA data logging, transmission, and activity/behavior inference system) for the testing and iterative refinement via laboratory simulators and test subjects.
This study is designed to extend previous work in the development of methods to automatically detect the timing of cocaine use from cardiac interbeat interval and physical activity data derived from wearable, unobtrusive mobile sensor technologies.
This study examines mechanisms of self-regulatory function via both passive physiological sensing and ecological momentary assessment (EMA) both within and outside of laboratory settings.
mCerebrum is a configurable software platform for mobile and wearable sensors. It provides support for reliable data collection from mobile and wearable sensors, and real-time processing of these data for sensor triggered just-in-time adaptive interventions.
Cerebral Cortex is the big data companion of mCerebrum designed to support population-scale data analysis, visualization, model development, and intervention design for mobile sensor data. It provides the ability to do machine learning model development on population scale data sets and provides interoperable interfaces for aggregation of diverse data sources.
Many smart home applications would like to know the room location of occupants, but tracking people in a convenient and practical way is notoriously difficult. We developed several new techniques for room location tracking by scanning a person's identity at the instant they cross the doorway threshold. One solution called the Doorjamb tracking system requires 1 sensor per doorway and can track the room location of multiple occupants with over 90% accuracy without requiring any RF transmitters or wearable tags, and without the use of privacy-invasive sensors such as cameras or microphones.