Optimizing non-invasive assessment of gastric activity: a fully automated channel-wise pipeline for multi-electrode electrogastrography

A fully automated channel-wise electrogastrography pipeline (GSPT) robustly recovers dominant-frequency gastric signals under controlled motion artifacts, improving signal-to-noise ratio.

• Why it matters: EGG adoption is limited by low signal-to-noise and the absence of standardized, objective channel-level artifact rejection and quality assessment.
• What they did: GSPT uses high-resolution time-frequency representations with entropy-based artifact detection and multi-strategy outlier detection, then computes channel and global gastric fingerprints (e.g., dominant frequency, Renyi entropy, Total Correlation).
• The result: In datasets with controlled motion artifacts, artifact identification precision/recall was comparable or better than existing methods, and gastric fingerprints were stable within subjects and across sessions.

AI summary of the abstract · may contain errors · not medical advice · read the paper.