A low complexity on-chip ECG data compression methodology targeting remote health-care applications

In this paper, we propose a novel low complexity on-chip ECG data compression methodology targeting remote health-care applications. This is to the best of our knowledge the first attempt for on-chip reliable data compression. The proposed methodology has been implemented targeting Application Specific Integrated Circuit platform at 1 MHz at V_dd 1.62 V for UMC 130 nm technology library with 16 bits system word-length. Furthermore the proposed methodology results in a faithful reconstruction which has been validated using MIT-BIH PTB-DB as well as our institute´s health repository IITH-DB. On an average about 90% compression is achieved with more than 83% R² statistics, 98% Cross Correlation and about 99% Regression between the original and the reconstructed data signifying the diagnostic accuracy. Subsequently the proposed methodology is capable of storing approximately 47 hrs of data in the same on-chip memory when compared to that of 5 hours of continuous data in the state of the art which would lead to enhanced diagnosis and prognosis in remote health-care.