Detecting aortic fiber architecture in ex-vivo arteries: A feasibility study with clinical 3T MRI

Abstract

Diffusion Tensor Imaging is a non-invasive imaging technique based on Magnetic Resonance Imaging that provides information on the tissue microstructure from the preferential direction of water molecules' diffusion. While the technique is widely used in neuroimaging, recent new applications were found for arterial tissue microstructure such as aorta and carotids. In the state of art, the Diffusion Tensor Imaging datasets for arterial tissues are usually acquired with ultra-high field scanners and no singular software for the processing of ex-vivo ring-like tissues is available. The present manuscript aims to demonstrate the application of a clinical magnetic resonance scanner to infer on fiber microstructure of ex-vivo arterial specimens. This was achieved by developing a custom workflow for the specific analysis of Diffusion Tensor Imaging data for arterial microstructures. First, a custom software platform was developed by including dedicated modules to perform the following processing pipeline: NIfTI conversion, eddy current correction, segmentation, estimation of diffusion tensor, diffusion parameters and fiber reconstruction and analysis. Then, a set of acquisitions was carried out on six fresh human aortic samples of ascending aorta by using a 3T clinical scanner (Philips Ingenia). The data were processed with both our Python-based custom workflow and other commercial software. The results obtained from the custom workflow were in agreement with the ones from the commercial software. Moreover, specific tools for tissue fibers visualizations and orientation analyses were added. In this study, the usefulness of the custom workflow for processing specific arterial Diffusion Tensor Imaging datasets was demonstrated. The efficacy of the processing pipeline was comparable with the other commercial software, with also the addition fiber analysis tools, specific for the vessel structure.

Keywords: Arterial tissue microstructures; Collagen fiber distribution; Diffusion tensor imaging; Image processing software; Magnetic resonance imaging.