Aorta Segmentation for Stent Simulation

TL;DR

This paper introduces a novel framework for virtual aortic stenting using CT scans, featuring a new segmentation method for lumen and outer wall detection validated on extensive data with high accuracy.

Contribution

The paper presents a new minimal closure tracking algorithm for outer wall segmentation and a comprehensive framework for virtual aortic stenting from CT data.

Findings

Achieved a Dice Similarity Coefficient of 90.67% on over 3000 MPR planes.

Validated the segmentation method on 50 CT angiography datasets.

Facilitates preoperative planning and device selection for aortic stenting.

Abstract

Simulation of arterial stenting procedures prior to intervention allows for appropriate device selection as well as highlights potential complications. To this end, we present a framework for facilitating virtual aortic stenting from a contrast computer tomography (CT) scan. More specifically, we present a method for both lumen and outer wall segmentation that may be employed in determining both the appropriateness of intervention as well as the selection and localization of the device. The more challenging recovery of the outer wall is based on a novel minimal closure tracking algorithm. Our aortic segmentation method has been validated on over 3000 multiplanar reformatting (MPR) planes from 50 CT angiography data sets yielding a Dice Similarity Coefficient (DSC) of 90.67%.