Groundtruth: A Matlab GUI for Artifact and Feature Identification in Physiological Signals

Area of Science:

• Physiological signal processing
• Biomedical engineering
• Wearable sensor technology

Background:

• Emerging wearable sensors require robust methods for performance assessment.
• Existing methods for artifact rejection and feature identification in physiological signals can be time-consuming.
• There is a need for user-friendly tools to evaluate new sensor technologies and algorithms.

Purpose of the Study:

• To introduce Groundtruth, a novel Matlab Graphical User Interface (GUI).
• To provide a platform for interactive identification of key features and artifacts in physiological signals.
• To facilitate the assessment of new sensor performance and automated algorithm evaluation.

Main Methods:

• Development of a customizable Matlab GUI (Groundtruth) for interactive physiological signal analysis.
• Implementation of simultaneous interactive marking of artifact regions and feature identification (e.g., respiration peaks, R-peaks).
• Application of the GUI to assess two simultaneously worn respiration sensors, including artifact removal and respiration rate computation.

Main Results:

• The Groundtruth GUI allows for efficient interactive marking of artifacts and features in physiological signals.
• Respiration rates computed using Groundtruth (post-artifact removal) showed consistent results between two simultaneously worn respiration sensors.
• Bland-Altman plots validated the consistency of respiration rate measurements after artifact removal.

Conclusions:

• Groundtruth offers a valuable, open-source tool for evaluating physiological sensor performance and algorithms.
• The GUI streamlines the process of data marking for artifact rejection and feature identification.
• Consistent results obtained from multiple sensors highlight the effectiveness of Groundtruth in physiological signal analysis.