An Automatic Technique for MRI Based Murine Abdominal Fat Measurement

Abstract

Because of the well-known relationship between obesity and high incidence of diseases, fat related research using mice models is being widely investigated in preclinical experiments. In the present study, we developed a technique to automatically measure mice abdominal adipose volume and determine the depot locations using Magnetic Resonance Imaging (MRI). Our technique includes an innovative method to detect fat tissues from MR images which not only utilizes the T1 weighted intensity information, but also takes advantage of the transverse relaxation time(T2) calculated from the multiple echo data. The technique contains both a fat optimized MRI imaging acquisition protocol that works well at 7T and a newly designed post processing methodology that can automatically accomplish the fat extraction and depot recognition without user intervention in the segmentation procedure. The post processing methodology has been integrated into easy-to-use software that we have made available via free download. The method was validated by comparing automated results with two independent manual analyses in 26 mice exhibiting different fat ratios from the obesity research project. The comparison confirms a close agreement between the results in total adipose tissue size and voxel-by-voxel overlaps.