Corpore Sano – Medicine

We are currently involved in several inter-disciplinary initiatives investigating novel technology solutions in the medical domain. Our main objective is to develop a general infrastructure where quantitative data can be collected continuously and objectively, in large cohorts, and over long time-spans. We are developing and evaluating efficient solutions on how to measure, monitor, collect, store, curate, access, combine, and process a large variety of existing and emerging fitness/health sensor streams and other relevant data sources. We are in particular investigating how a sensor-rich, geo-distributed environment connected to soft real-time analytics (machine learning) software can provide the basis for personalized health intervention technologies.

Security and privacy for sensitive data throughout the infrastructure are key requirements – while still providing the option to securely use machine learning algorithms to extract patterns and to enable early prediction and personalized feedback based on fitness/health sensor signals.

New types of (machine learning) analytics can potentially be carried out in near real-time on empirical measurements, which will complement existing epidemiological research methods. We hypothesize that new medical insights can be found at the intersection of Big Data technologies and epidemiology, a field of medicine building on large data registers and analysis where Norway already is excellently positioned internationally.

We are also involved in performing research and development of end-to-end infrastructures for non-invasive, wireless capsule endoscopy of the entire digestive system. Focus is on developing algorithmic systems that target detection of different irregularities in the human digestive system. This includes detection of colon polyps, Chron’s disease or Colorectal cancer using video object tracking, object detection, machine learning or other relevant mechanisms. A series of prototype systems (EIR) (OpenSea) have been developed, and we have made public several gastrointestinal data sets (KVASIR, Nerthus).