Effective high compression of ECG signals at low level distortion

Abstract

An effective method for compression of ECG signals, which falls within the transform lossy compression category, is proposed. The transformation is realized by a fast wavelet transform. The effectiveness of the approach, in relation to the simplicity and speed of its implementation, is a consequence of the efficient storage of the outputs of the algorithm which is realized in compressed Hierarchical Data Format. The compression performance is tested on the MIT-BIH Arrhythmia database producing compression results which largely improve upon recently reported benchmarks on the same database. For a distortion corresponding to a percentage root-mean-square difference (PRD) of 0.53, in mean value, the achieved average compression ratio is 23.17 with quality score of 43.93. For a mean value of PRD up to 1.71 the compression ratio increases up to 62.5. The compression of a 30 min record is realized in an average time of 0.14 s. The insignificant delay for the compression process, together with the high compression ratio achieved at low level distortion and the negligible time for the signal recovery, uphold the suitability of the technique for supporting distant clinical health care.

Wavelet based dictionaries for dimensionality reduction of ECG signals

“…In that sense the adopted strategy produces results which significantly improve upon the state of the art for the full MIT-BIH Arrhythmia data set [40][41][42][43]. In regard to CR the results are close to those recently reported in [37]. In that publication the compression results are obtained with an equivalent strategy as that described in Sec.…”

“…3, but applying a FWT on the full signal instead of using a piecewise model as proposed here. With respect to processing time the approach in [37] is far more efficient though. Indeed, with the same equipment the construction of Table 2 However, the central aim of the present work is to accomplish dimensionality reduction.…”

“…The 4th column gives the mean value CR and the fifth column the corresponding standard deviation. For further improvement in CR we have added an entropy coding step to represent the strings described Fast Wavelet Transform (FWT) is applied on the whole record, as proposed in [37].…”

“…In [42] the lossless compression is realized using Huffman coding while [41] applies symbol substitution. Comparisons are also produced with the results arising when a FWT is applied on the full record [37]. Table 3 compares CRs obtain with CDF97 D and CDF53 D for values of PRD in the range [0.31 0.81], Table 4 for values in the range [0.91 1.06], and Table 5 in the range [0.…”

“…These parameters are set in order to reproduce the same values of PRD as in the publications we are comparing with. Table 3: Comparison between the average compression performance of the proposed method and the methods in [40], [41], [42] and [37], for values of PRD in the range [0. Table 4: Comparison between the average compression performance of the proposed method and the methods in [41], [42], [43] and [37], for values of PRD in the range […”