Real-time ECG data transmission with Wavelet Packet Decomposition over wireless networks

With the increase of demand for remote cardiac monitoring for elderly patients, medical service providers are facing a challenging problem: how to efficiently transmit continuous holter ECG data over often bandwidth constrained wireless network so that transmitted data is quality-wise good enough for doctors or automated computer systems to detect cardiovascular conditions in almost real-time? This paper proposes a novel solution to address this challenge. One obvious solution to efficiently transmit ECG data is to use existing ECG compression applications that can reduce the data size to a great extent. However, holter ECG data even from a single patient could be huge, and when data of several patients is transmitted from the same facility, compression alone may become ineffective due to congestion causing data loss of compressed data. In this paper, we propose a transmission system which first classifies ECG data into low, medium and high frequency components using Wavelet Packet Decomposition (WPD), and then uses priority queuing system where low frequency components containing the most significant features of cardiovascular conditions can be given highest priority before being transferred over bandwidth-constrained wireless networks. Performance results show that this scheme can be hugely beneficial.