Motion Artifact Data to Facilitate Bioelectric Signal Quality Analysis Research
Jonathan Kulpa · 2023
Bioelectric signal quality analysis is important to address the challenges associated with contaminants, including motion artifacts.However, the limited availability of motion artifact data poses a challenge in developing and evaluating new tools (e.g., biases due to signal reuse).This research expands motion artifact data through two approaches.First, we deployed a Motion Artifact Signal Generation Toolkit to synthesize motion artifacts using Autoregressive, Markov Chain, and Recurrent Neural Network models.We extend model validation to non-cyclical motion artifacts.Second, we estimate motion artifacts in longterm ECG recordings using a template subtraction method, creating a motion artifact database of 84 signals.Additionally, we explore time-series segmentation of motion artifacts, leveraging a clustering pipeline, including k-means clustering, to partition longterm recordings based on signal statistics and attributes.These contributions greatly expand the public availability of motion artifacts for biomedical signal quality analysis researchers to use worldwide.vi 5.2.7 Post-Processing ......