Authors:
Axel Helmer
1
;
Riana Deparade
2
;
Friedrich Kretschmer
3
;
Okko Lohmann
1
;
Andreas Hein
1
;
Michael Marschollek
4
and
Uwe Tegtbur
2
Affiliations:
1
OFFIS Institute for Information Technology, Germany
;
2
Institute of Sports Medicine, Germany
;
3
University of Oldenburg, Germany
;
4
University of Braunschweig - Institute of Technology and Hannover Medical School, Germany
Keyword(s):
Modeling, Heart Rate, prediction, cardiopulmonary rehabilitation
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Biomedical Engineering
;
Business Analytics
;
Cloud Computing
;
Data Engineering
;
Data Mining
;
Databases and Information Systems Integration
;
Datamining
;
Decision Support Systems
;
e-Health
;
Enterprise Information Systems
;
Health Information Systems
;
Physiological Modeling
;
Platforms and Applications
;
Sensor Networks
;
Signal Processing
;
Soft Computing
;
Telemedicine
Abstract:
Chronic obstructive pulmonary disease (COPD) and coronary artery disease are severe diseases with increasing prevalence. They cause dyspnoea, physical inactivity, skeletal muscle atrophy and are associated with high costs in health systems worldwide. Physical training has many positive effects on the health state and quality of life of these patients. Heart Rate (HR) is an important parameter that helps physicians and (tele-) rehabilitation systems to assess and control exercise training intensity and to ensure the patients’ safety during the training. On the basis of 668 training sessions (325 F, 343 M), demographic information and weather data, we created a model that predicts the training HR for these patients. To allow prediction in different use cases, we designed five application scenarios. We used a stepwise regression to build a linear model and performed a cross validation on the resulting model. The results show that age, load, gender and former HR values are important pred
ictors, whereas weather data and blood pressure just have minor influence. The prediction accuracy varies with a median root mean square error (RMSE) of ≈11 in scenario one up to ≈3.2 in scenario four and should therefore be precise enough for the application scenarios mentioned above.
(More)