PI's: Prof. Dr. H.E. Wichmann, Prof. Dr. Alexander Horsch    
PhD Candidates: Lukas Gorzelniak, Dipl.-Inf. Med.; André Dias, M.Sc.

The Co-operative Health Research in the Region of Augsburg (KORA) project, directed by Prof. Wichmann, has systematically collected data from a large population-based cohort of some 18,000 subjects living in the city of Augsburg and the surrounding counties. This treasure of a data collection (20+ years) enabled systematic and consistent long-term research.Today, with the emergence of biosensors, the set of parameters acquired from the cohort can be enhanced, by their capability of measuring a large scope of body parameters in an objective way.  In the proposed Ph.D. project two such parameters are addressed: physical activity and heart function.

For the measurement of physical activity, physiological acceleration sensors are available. They provide a means for the measurement of physical activity in the common environment of the human subject which is under monitoring. Studies have shown that patients with Chronic Obstructive Pulmonary disease (COPD) and little daily physical activity are at higher risk of re-hospitalization and shorter survival time compared to those with high level of physical activity. Currently, the common approach for quantifying the physical activity of COPD patients is based on questionnaires or diaries, analysis of video recordings, or data from movement sensors.

The questionnaire approach has serious drawbacks. For example, it has been shown that COPD patients tend to overestimate the duration of walking periods when asked to document these in a diary. It was also found that with questionnaires high physical activity was better reflected than moderate and low activity which is typical for subjects with COPD. For the measurement of heart activity, simple and more sophisticated devices are available.The simple ones typically measure an average heart rate on e.g. the basis of 1-minute temporal window. The more sophisticated ones are able to measure the duration of the individual heart cycles (R-R). The potential of these heart activity parameters for the assessment of COPD conditions is not well understood.

In particular, the analysis of complex patterns in the variability of heart beat lengths and their correlation with other clinical data requires further research. The principal research question of the Ph.D. project is: Can classes representing different degrees of severity of COPD be discriminated by activity level and heart conditions measured with sensors? We hypothesize that there is a correlation between sensor data and the presence or severity of COPD conditions. Sub-questions include the investigation of possible correlations between: measured heart rate variability and COPD condition; measured physical activity and questionnaire-based self-assessment; measured level of physical activity and lung function; multi-modal sensor data features and COPD condition. The sensors are used on two groups within 10 measurement periods on in total 320 subjects from the KORA cohort, and the measured data are compared with the KORA data. In each measurement period, 16 COPD and 16 non-COPD subjects are equipped with the three different sensors. The measured data are processed using various signal processing and data analysis methods and correlation analyses are performed.

The study participants are a sub-cohort of the 1000 subjects from the KORA cohort around the city of Augsburg, 65 to 89 years old, who are going to be examined in the study center. Sensors and clinical data shall in depth be analyzed and adequate models for COPD conditions shall be created in the course of the study. The Ph.D. project will build on the infrastructure established as part of the KORA-Age project (Wichmann et al. 2008).It is envisioned as a component to the ongoing examinations of 1000 participants of the KORA-Age project, who will receive a detailed clinical examination to assess the determinants and consequences of multi-morbidity in subjects aged 65 and older.

By answering its specific research questions, this Ph.D. project is expected to create highly relevant evidence on how to monitor and manage COPD patients living independently in their own homes, by use of suitable biosensors. This is the first effort to study the opportunities that biosensors (other types of sensors will follow) can offer for studies within the KORA cohort and its 18,000 members, on a potentially large variety of specific problems. The project is based at the Chair of Medical Informatics, TUM Medical Center "Rechts der Isar", in cooperation with the Institute of Epidemiology at Helmholtz Center Munich.