GIQuant - Motility analysis
The dynamic series seen below can be acquired on any modern MRI scanner providing an excellent physiological view of the small bowel peristalsing. This process is essential for mixing our food and moving it along through the gastrointestinal (GI) tract. MRI is safe, well tolerated by patients and non-invasive making it an excellent modality for physiological investigations of the bowel.
The key limitation lies in the analysis of these images which until recently has been done subjectively by visually assessing the raw data or with manual annotation of single bowel loops which is time consuming and unsuitable for clinical practice.
GIQuant is a image processing software developed to facilitate our analysis of dynamic image series using image registration. The registration process warps each image in the time-series to a target 'reference' image producing a deformation field for each time point. These deformation fields can be visualised below as a grid with higher changes in deformation apparent in the small bowel compared to the liver. These deformation fields are exceptionally useful for motility analysis to both automatically propagate regions of interest (ROIs) and to produce parametric motility maps.
A parametric map is a form of quantitative representation of some aspect of the raw data which in this case is motility in the small bowel. In the images below we can see the reference image on the left and the motility map on the right. The reference image is one of the frames taken from the time series to which all the other time points were registered and serves as a useful guide for the identification of key anatomical landmarks. The motility map shares a 1:1 relationship with the reference image making ROI analysis in any DICOM browser easy with the numerical value of the map representing the magnitude of the motile actions at that location. We summarise motility by taking the standard deviation of the Jacobian determinant of the deformation field. This metric has been widely published but is quite abstract, for more information on what the SD Jacobian metric is, click here.
For intuitive interpretation of the motility maps, the reference image and map can be fused to allow robust scientific investigation and convenient clinical evaluation based on quantitative information.