Sensor vs. Human: Comparing Sensor Based State Monitoring with Questionnaire Based Self-Assessment in Bipolar Disorder Patients

Agnes Grünerbl, Gernot Bahle, Raphaela Banzer, Stefan Öhler, Christian Haring, Paul Lukowicz

In: Proceedings of the 18th International Symposium on Wearable Computers. IEEE International Symposium on Wearable Computers (ISWC-2014) September 13-17 Seatlle Washington United States ACM 9/2014.


We compare the performance of a smart phone based state and state change detection system to self-assessment and show that the automatic detection is much closer to the objective psychiatric diagnosis. Our work is based on a large, real life dataset collected with 9 real patients with a total of 800 days of data. It consists of smart phone sensor data, a daily self-assessment questionnaire filled out by the patients and is validated by standardized psychiatric scale tests.

German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz