HRV (Heart rate variability, which has a function of prediction for cardiovascular disease) contains a wealth of medical information, rapid extraction and procession of these signals will bring an important meaning for the prevention of heart diseases. Physionet open source project provides a good platform for the research and development of HRV, which also provides demonstration tools for the calculation of HRV. The characteristics of medical signal are real-time and have large volume of data. Conventional serial methods are difficult to meet the requirements of biomedicine, and the parallel method based on multi-core CPU is larger communication overhead. In this paper, we designed some parallel algorithms for the calculation of HRV in time-domain based on the strategy of parallel reduction, compared and analyzed the various optimization methods, and received the highest 38 times speedup compared with serial method.
CITATION STYLE
Wang, J., Chen, W., & Hou, G. (2015). Parallel computing method for HRV time-domain based on GPU. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9529, pp. 434–443). Springer Verlag. https://doi.org/10.1007/978-3-319-27122-4_30
Mendeley helps you to discover research relevant for your work.