Running Head : Physiological classification of functional bowel disorders

24 Introduction. We have previously developed an original method to evaluate 25 small bowel motor function based on computer vision analysis of endoluminal 26 images obtained by capsule endoscopy. Our aim was to demonstrate intestinal 27 motor abnormalities in patients with functional bowel disorders by endoluminal 28 vision analysis. Methods. Patients with functional bowel disorders (n=205) and 29 healthy subjects (n=136) ingested the endoscopic capsule (Pillcam-SB2, Given30 Imaging) after overnight fast and 45 min after gastric exit of the capsule a liquid 31 meal (300 ml, 1 kcal/ml) was administered. Endoluminal image analysis was 32 performed by computer vision and machine learning techniques to define the 33 normal range and to identify clusters of abnormal function. After training the 34 algorithm, 196 patients and 48 healthy subjects, completely naïve, were used 35 as test set. Results. In the test set, 51 patients (26%) were detected outside the 36 normal range (p<0.001 vs 3 healthy subjects) and clustered into hypo and 37 hyperdynamic subgroups As compared to healthy subjects, patients with 38 hypodynamic behavior (n=38) exhibited less luminal closure sequences (41±2 39 % of the recording time vs 61±2 %; p<0.001) and more static sequences (38±3 40 % vs 20±2 %; p<0.001); in contrast, patients with hyperdynamic behavior 41 (n=13) had an increased proportion of luminal closure sequences (73±4 % vs 42 61±2 %; p=0.029) and more high motion sequences (3±1 % vs 0.5±0.1 %; 43 p<0.001). Conclusion. Applying an original methodology we have developed a 44 novel classification of functional gut disorders based on objective, physiological 45 criteria of small bowel function. 46