A Case Study using Simple Moving Average Filters to accomplish ECG denoising on an FPGA

Electrocardiography (ECG) results are essential in identifying a variety of cardiac disorders. The ECG is often conducted with accurate electronic equipment because the heart's pulses can alter dramatically and noise can impair decision accuracy. Exact measurements are essential when using cutting-edge software to extract data from an ECG dataset and make choices about various types of cardiac disorders. Artifacts and noise can corrupt even the most precise data. While being acquired and processed, the ECG signal may become contaminated by high- and low-frequency noise such as electromyogram noise, additive white Gaussian noise, and power line interference (baseline wandering). ECG signals must be denoised since interference could result in incorrect interpretation. On an Altera DE2-115 board, the results of various pipelining and cascaded-integrator-comb (CIC) design approaches are explored. Pipelining in the FIR portion (comb) and the recursive part lowers latency, but because the division by 8 happens at the end, "CIC_dive_pipe" performs better than "CIC_divm_pipe" (642.67 MHz) at the expense of more LEs (66).