A new method to measure cross sectional area of vessels in MRI image and its application in stenosis detection

A new method is introduced in this paper to determine the best means by which to measure the cross sectional area in magnetic resonance images. Images are collected using a three dimensional gradient echo imaging sequence. A dynamic programming based segmentation algorithm is employed to determine the vessel cross section on a slice-by-slice basis. We then estimate the segmentation accuracy as a function of vessel size, sampling matrix, acquisition time and post processing method. We show that there is an optimal value at which the segmentation error is smallest for a fixed set of the aforementioned parameters and that this value scales with the ratio of the vessel diameter to the resolution. Based on the results, it is finally possible to describe how best to collect the data in MRI to be able to determine the vessel´s cross sectional area within a certain accuracy and precision, which is a key missing linking of image processing and image acquisition in MRI field.