VCG signal compression based on singular value decomposition

Vectorcardiography (VCG) has emerged as a physiological monitoring technique to effectively represent vital signals through the electrical activity of projected signals on spatial planes. However, utilising VCG data to monitor heart disease produces enormous volumes of clinical data. Therefore, VCG data must be compressed for transmission and storage. This work presents a simple yet efficient technique for compressing VCG signals, using singular value decomposition (SVD). The SVD transformation effectively extracts the considerable signal and robust interbeat correlation in VCG cycles, leading to optimal pattern decomposition and reduced redundant data. The effectiveness of the proposed compression approach is assessed using 10 different VCG data based on different fidelity metrics, such as CR, PSNR, PRD, SSIM, MSE, and SNR. The proposed compression technique achieves an average CR of 6.55, PSNR of 38.13 dB, PRD of 4.90, SSIM of 0.966, SNR of 28.09 and MSE, 2.61 × 10−4. The results illustrate that the proposed compression technique yields a good CR for VCG data while maintaining significant diagnostic information.