Affordable Access

deepdyve-link
Publisher Website

Accuracy of collagen fibre estimation under noise using directional MR imaging.

Authors
  • Brujic, Djordje1
  • Chappell, Karyn E2
  • Ristic, Mihailo3
  • 1 Mechanical Engineering Department, Imperial College London, London, UK.
  • 2 MSK Lab, Department of Surgery and Cancer, Imperial College London, UK.
  • 3 Mechanical Engineering Department, Imperial College London, London, UK. Electronic address: [email protected]
Type
Published Article
Journal
Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society
Publication Date
Dec 01, 2020
Volume
86
Pages
101796–101796
Identifiers
DOI: 10.1016/j.compmedimag.2020.101796
PMID: 33069034
Source
Medline
Keywords
Language
English
License
Unknown

Abstract

In tissues containing significant amounts of organised collagen, such as tendons, ligaments, menisci and articular cartilage, MR imaging exhibits a strong signal intensity variation caused by the angle between the collagen fibres and the magnetic field. By obtaining scans at different field orientations it is possible to determine the unknown fibre orientations and to deduce the underlying tissue microstructure. Our previous work demonstrated how this method can detect ligament injuries and maturity-related changes in collagen fibre structures. Practical application in human diagnostics will demand minimisation of scanning time and likely use of open low-field scanners that can allow re-orienting of the main field. This paper analyses the performance of collage fibre estimation for various image SNR values, and in relation to key parameters including number of scanning directions and parameters of the reconstruction algorithm. The analysis involved Monte Carlo simulation studies which provided benchmark performance measures, and studies using MR images of caprine knee samples with increasing levels of synthetic added noise. Tractography plots in the form of streamlines were performed, and an Alignment Index (AI) was employed as a measure of the detected orientation distribution. The results are highly encouraging, showing high accuracy and robustness even for low image SNR values. Copyright © 2020 The Authors. Published by Elsevier Ltd.. All rights reserved.

Report this publication

Statistics

Seen <100 times