Extrakce tepové frekvence z hrudního pásu

Abstract

This bachelor thesis focuses on the extraction of heart rate from the breathing curve. A chest belt, equipped with a piezo transducer, is used to measure the mechanical activity of the heart and the breathing curve. The aim of the thesis is to summarize the theoretical principles of ballistocardiography, investigate signal processing methods and apply custom algorithms for heart rate detection. In the practical part, experimental measurements of different respiratory frequencies and application of the proposed algorithms were performed. Statistical analysis of the data provided a deeper understanding of the results and information for further optimization of the detection process. Interquartile variance of differences in BPM detections and hypothesis testing were key tools to assess the success and reliability of the detection methods. The results suggest that the proprietary algorithm achieves higher reliability compared to the other methods, except in post-physical activity situations where adaptive thresholding appears to be more effective. Suggested improvements to the methods include signal averaging and improved analogue preprocessing to further increase detection accuracy.