Third-order tensor decomposition for Multilead ECG

What is it about?

Multilead ECG data has three types of correlations. The 3-array representation of MECG data helps exploit all three types of correlations. This reduces data dimension significantly.

Featured Image

Why is it important?

With ever-increasing types of cardiac ailments, it is necessary to store the patient data for future analysis. Hence, data dimension reduction always remains an ambitious problem for engineers. We have proposed a novel method to represent the multilead ECG data in third-order tensor form. Our findings show that the data size can be reduced by a factor of 45 times, and outperforms the state-of-the-art techniques.