ECG signal gives the information of heart beat rate of a person. When a patient needs an ECG, which is a non-surgical technique to measure and record the electrical activity of the heart, a skilled cardiologist may not always be on site. Consequently, a general practitioner or paramedical team visiting the patient's location must use a particular kind of automated ECG analysis in order to obtain the patient's electrocardiogram. Automated ECG analysis is required. The optimal methodology for automated analysis of multi-channel ECG signals is finally provided in this paper. Artifacts and noise in multi-channel ECG readings can influence the diagnosis. By employing a neural network to determine the dynamic cutoff frequency parameter from noisy ECG signals and the particle swarm improvement approach (RBFNN-PSO), some researchers were able to reduce noise from the signals. However, PSO only offers Swarm, and reaction times are lengthy. A better RBFNN-PSO technique known as Radial Basis Function Neural Network with Multi Swarm Optimisation (RBFNN-MSO) has been proposed in order to get over these drawbacks. Lastly, the RBFNN-MSO technique, which is used with digital low-frequency filters for impulse response (FIR), establishes the cutoff frequency parameter.