Vectorcardiogram signal compression: A hybrid approach using discrete wavelet transform and singular vector sparse reconstruction

Vectorcardiography (VCG) records the heart's electrical activity in three dimensions: frontal, transverse, and sagittal for accurate cardiac diagnosis and monitoring. Continuous cardiac monitoring generates vast VCG data daily. The large amount of VCG data poses challenges for storage, dataset management, and transmission, particularly in remote areas with limited resources. The main objective of this research is to develop an efficient VCG data compression technique to streamline the handling and storage of cardiac data, reducing storage space, bandwidth use, and enhancing data transmission speed.

This work proposes a VCG data compression technique combining Discrete Wavelet Transform (DWT) and Singular Vector Sparse Reconstruction (SVSR) to reduce data size, while preserving vital cardiac information. The process consists of two stages: compression and reconstruction. In compression, the VCG signal undergoes decomposition using the Discrete Wavelet Transform with Haar wavelet, extracting low and high-frequency components. Moreover, the compression efficiency is enhanced by applying SVSR in each subband. In the reconstruction phase, the compressed VCG signal is restored by performing interpolation and the inverse discrete wavelet transform (IDWT), improving overall reconstruction performance.

The results have been tested using PTB diagnostic ECG databases. The effectiveness of the developed compression method is examined quantitatively and qualitatively, in terms of the following metrics; such as compression ratio, signal-to-noise ratio, peak signal-to-noise ratio, percent root mean square difference (PRD) and structural similarity index. According to the performance metrics results, the proposed method achieves 55.67 % higher CR, and 57.12 % improved PRD as compared to existing method.

The proposed method is efficient and adaptable for compressing different VCG signals, offering control over the quality of reconstructed data. Simulated results illustrate that the proposed compression method is a very promising solution for data storage and tele-transmission applications.