The paper presents a K-NN based interpolation for handling artifacts in IBI signal for HRV analysis.
The algorithm detects artifacts in IBI signals and interpolates them with artifacts free consecutive before and after parts of the corrupted area.
The detection is performed applying windowing technique based on mean and a threshold values.
The interpolation is done on the basis of the K-NN algorithm which allows us to use closet data points for interpolation for an individual.
Though the data points before and after the corrupted area might not be exactly the
same but for a specific individual it persists some evidence of the underlying characteristics of the signal.
Therefore, we prefer K-NN as it considers nearest neighbors.
The experimental work illustrates that the algorithm is able to detect and interpolate artifacts from IBI signals.
Though the proposed algorithm in some cases provides results that is not better than some other techniques involved in the experiments it provides a better solution
where manual interference, domain knowledge, loss of data points and interpolation of the whole part of signal would be a problem.