Method for artefact detection and removal in heart rate signals measured during physical exercise

Monitoring the training of athletes is essential for success in sports. Information that is derived from measured data during the training is only reliable if the quality of the data is high. Therefore artefact detection and removal are essential. In this paper typical artefacts in heart rate measurements during football field-trainings are described. An algorithm to automatically detect artefacts and an algorithm to pre-process them are also presented. The results show that with the proposed pre-processing method the percentage of artefacts can be significantly (p < 0.01) reduced from 9.0 to 4.4 %. This corresponds to a total period of 10 hours of unreliable data on a total of 126 hours that have been taken out. As a result, more high quality data is available for monitoring training. The developed methods are generic and can be used in many applications where accurate heart rate monitoring is crucial.status: publishe