Mobile devices and sensors are profoundly changing the way we create, consume, and share information. Health aficionados and citizens are increasingly using sensors and devices to track sleep, food, activity, and other physiological observations [2]. This trend is moving toward a paradigm shift of reactive medicine to proactive and preventive medicine. We have built a system called kHealth that is evolving into a proactive and continuous monitoring of observations from Physical, Cyber, and Social domains [1] for health and well-being. kHealth is a knowledge-enabled semantic platform to enhance decision making and improve health, fitness, and well-being. It supports contextual (e.g., condition specific) annotation, integration, and interpretation of sensor and mobile data from individuals using deep domain (e.g., disease) specific knowledge base.

Research Areas

Semantic Perception: There are increasingly affordable sensors for monitoring activity levels, food intake, sleep quality, physiological observations (Blood pressure, Heart rate, Weight, Temperature). These sensors generate massive amounts of data and analyzing these observations manually is not feasible. We need intelligent ways of processing these observations to gain insights. Research on Semantic Perception [4, 5] deals with moving from massive, unintuitive raw sensor observations to compact, intelligible abstractions.

Understanding EMR/PHR Records: EMR (Electronic Medical Records) and PHR (Personal Health Records) are increasingly being adopted for documenting patient visits. They capture the digital footprint representative of the health of individuals with fine details such as chronic conditions, prescribed medications, demographic details, and temporal aspects of disease, and medication outcomes. The meaningful use of these documents is being restricted by the unstructured nature of the documents. We are building techniques to improve the machine understanding of these documents. This step helps us in improving the richness of contextual information about the patient.

Risk Assessment: While there are massive, heterogeneous, and multi-modal observations from sensors, mobile devices, personal observations, and EMR/PHRs, we believe that providing actionable information to an individual is very valuable. For providing actionable information, we need to assess the health condition of the individual by collectively analyzing observations from Physical, Cyber, and Social worlds. Research on risk assessment is toward building contextual and personalized risk profiles, which in turn can be used for recommending actions to the individual.

Research Applications


In collaboration with Dr. Maninder Kalra since December 2012, we are working on algorithms to analyze, synthesize, and recommend actions to pediatric asthma patients. Please look at the kHealth for asthma project page for more details on this:


In collaboration with Dr. Larry Lawhorne (Chair and Professor of Geriatrics, Boonshoft School Of Medicine) we are developing algorithms on analyzing patient behavior from low-cost sensors and mobile devices for measuring dementia behavior episodes as well as caregiver stress.  Please look at the kHealth for dementia project page for more details on this:


In collaboration with Dr. Sangeeta Agrawal (Chief, Division of Gastroenterology, Dayton VA Medical Center) we are developing algorithms on analyzing observations from low-cost sensors and mobile devices for reducing re-admissions of patients with liver cirrhosis.

ADHF (Acute Decompensated Heart Failure)

In collaboration with Dr. William Abraham (Director, Division of Cardiovascular Medicine), Wexner Medical Center, Ohio State University, we have developed a mobile sensor application toward Reducing preventable readmissions for ADHF patients.




KHealth Video Introduction :

ADHF App Walkthrough :

Asthma App Walkthrough :





  1. Amit Sheth, Pramod Anantharam, Cory Henson, 'Physical-Cyber-Social Computing: An Early 21st Century Approach,' IEEE Intelligent Systems, pp. 79-82, Jan./Feb. 2013
  2. eHealth Sensor Platform
  3. Wolf, G., A. Carmichael, and K. Kelly. "The quantified self." TED html (2010).
  4. Cory Henson, Krishnaprasad Thirunarayan, and Amit Sheth, 'An Efficient Bit Vector Approach to Semantics-based Machine Perception in Resource-Constrained Devices,' In: Proceedings of 11th International Semantic Web Conference (ISWC 2012), Boston, Massachusetts, USA, November 11-15, 2012.
  5. Cory Henson, Amit Sheth, Krishnaprasad Thirunarayan, 'Semantic Perception: Converting Sensory Observations to Abstractions,' IEEE Internet Computing, vol. 16, no. 2, pp. 26-34, Mar./Apr. 2012, doi:10.1109/MIC.2012.20