COMPUTERIZED ANALYSIS OF AMBULATORY LONG-TERM SMALL-BOWEL MANOMETRY

Background: Ambulatory long-term manometry is increasingly being used to study small-bowel motility. This study aimed to develop computer-ai ded data analysis including the elimination of artefacts, identificati on of individual phasic contractions, and analysis of aboral propagati on. Methods: Data processing included low-pass filtering, base-line ad aptation, cross-comparison of channels, and application of threshold v alues for contraction parameters. Automated analysis was validated by a visual reference standard.Results: Artefacts were related to cardiov ascular and respiratory activity, changes in body posture, and contrac tions of the abdominal wall. Automated recognition of contractions rea ched a sensitivity of 92% and a positive predictive value of 88% compa red with the visual standard. Mean contraction amplitude and duration of computer analysis were 96% and 93%, respectively, of the visually o btained values. Propagation analysis under ambulatory conditions showe d good agreement with previous results by stationary recordings. Concl usions: Computerized analysis provided valid and reproducible data on small-bowel phasic contractile events and propagative activity by digi tal long-term manometry.