Blanchett T, Kember G C, Fenton G A
Department of Applied Mathematics, Technical University of Nova Scotia, Halifax, Canada.
IEEE Trans Biomed Eng. 1998 Jul;45(7):942-5. doi: 10.1109/10.686803.
An electrocardiogram (ECG) compression algorithm based on a combination of the Karhunen-Loeve transform (KLT) and multirate sampling is introduced. The use of multirate sampling reduces KLT computational times to those reported for wavelet-packet-based compression techniques. A beat-by-beat quality controlled compression criterion is shown to be necessary to ensure clinically adequate reconstruction of each beat. The resulting quality controlled algorithm efficiently achieves compression rates of approximately 30-40:1 for the MIT-BIH database.