Meeting Banner
Abstract #0875

Non-Invasive MRI-Based 3D Volumetric Serial Assessments of Physiologic Large Intestine Gas - Proof of Principle

John Butler1, Jodi Miller1,2, Harry Marshall1,2, Ally Silavi1, John Patrick1,2, William Pavlosky3, Gregor Reid1,4, Don Taves3, Jamie Gregor5, Khaleel Sultan5, Deanna Carlsen6, Artem Khlebnikov6, Denis Guyonnet7, Sophie Legrain-Raspaud7, Frank S. Prato1,2, R. Terry Thompson1,2, Robert Z. Stodilka1,2

1Lawson Health Research Institute, London, Ontario, Canada; 2Medical Biophysics, University of Western Ontario, London, Ontario, Canada; 3Nuclear Medicine, St. Joseph's Hospital, London, Ontario, Canada; 4Microbiology, University of Western Ontario, London, Ontario, Canada; 5Gastroenterology, London Health Sciences Center, London, Ontario, Canada; 6The Dannon Company, White Plains, NY, United States; 7Danone Research, Palaiseau, Cedex, France


This study uses a single-shot MRI technique to provide a 3D dataset for semi-automatic image segmentation to determine large bowel gas volume. An axial interleaved single shot HASTE breath-hold sequence, acquired in isocenter mode at three table positions was used. Three sets of images were acquired in five subjects on a restricted diet, over a day for two consecutive days. Prior to the last two scans on the 2nd day the subjects were given a lactulose challenge. The total gas volume in the large bowel was calculated using an operator interactive segmentation routine to extract the voxels containing gas. The measured gas volumes were consisted in the same subject in Day 1 and 2, but there was larger inter subject variability in the gas volumes. The lactulose did not have a significant effect on gas volumes.