Validation of New GIMA Biomarker Signature of Endometriosis- Interim Data

Objective: Validate diagnostic accuracy of new unique biomarker, Gastrointestinal Myoelectrical Activity (GIMA),detected by Electroviscerography (EVG) with Ai-derived disease threshold score calculation to non-invasively diagnoseendometriosis.Design: Multicenter prospective blinded trial.Setting: Women’s healthcare center.Population of Sample: 165 patients with and without endometriosis diagnosis.Methods: Initial 50 patients meeting inclusion criteria in 165-patients multicenter prospective GIMA biomarker trial wereselected for interim analysis. Study population included women 27 years-55 years old, 25 with diagnosis of endometriosisand 25 non-endometriosis controls. Clinical and GIMA data were collected between February 2007 and September 2017,at all harvesting time points and frequency bands using EVG. Ai-derived threshold score calculations used Area UnderThe Curve (AUC), age and standardized pain scores variables.Main Outcome Measures: Specificity, sensitivity, NPV, PPV and predictive probability or C-statistic from logisticalregression analyses of all AUC frequency and time points.Results: Non-endometriosis versus endometriosis cohort interim analysis differed significantly (p<0.001) for median(IQR), AUC values, and percent frequency power distribution at baseline, (10, 20, and 30) minute post water-load atfrequency ranges (15-20, 30-40, 40-50 and 50-60) cpm. GIMA threshold scoring revealed sensitivity and PPV of 96%,specificity and NPV of 96% and C-statistic of 100%. Ai-derived GIMA biomarkers threshold scoring predicted 25/25subjects positive and negative for endometriosis, with surgical confirmation. Hormonal therapy, surgical stage, age norpain score affected diagnostic accuracy.Conclusion: EVG GIMA biomarker data with Ai-derived threshold scoring accurately distinguished participants with andwithout endometriosis. This interim analysis supports continued investigation of GIMA biomarkers to diagnoseendometriosis.