Free-Water Correction in Diffusion MRI: A Reliable and Robust Learning Approach

Weninger L, Koppers S, Na CH, Juetten K, Merhof D (2020)

Publication Type: Book chapter / Article in edited volumes

Publication year: 2020

Publisher: Springer Science and Business Media Deutschland GmbH

Series: Mathematics and Visualization

Pages Range: 91-99

DOI: 10.1007/978-3-030-52893-5_8

Abstract

In clinical settings, diffusion MRI can be used for extracting biomarkers such as fractional anisotropy or for revealing brain connectivity based on fiber tractography. Both are impacted by the free-water partial volume effect that arises at the border of cerebrospinal fluid or in presence of vasogenic edema. Hence, in order to robustly track white matter fibers close to cerebrospinal fluid and in presence of edema, or to extract consistent biomarkers in these cases, the diffusion MRI signal needs to be corrected for partial volume effects. We present a novel method that reproducibly infers plausible free-water volumes across different diffusion MRI acquisition schemes. Based on simulated data closely following the individual characteristics of each measurement, a neural network is trained on synthetic groundtruth data. According to our evaluation, this methodology produces more consistent and more plausible results than previous approaches.

Involved external institutions

Rheinisch-Westfälische Technische Hochschule (RWTH) Aachen DE Germany (DE) Universitätsklinikum Aachen (UKA) DE Germany (DE)

How to cite

APA:

Weninger, L., Koppers, S., Na, C.H., Juetten, K., & Merhof, D. (2020). Free-Water Correction in Diffusion MRI: A Reliable and Robust Learning Approach. In (pp. 91-99). Springer Science and Business Media Deutschland GmbH.

MLA:

Weninger, Leon, et al. "Free-Water Correction in Diffusion MRI: A Reliable and Robust Learning Approach." Springer Science and Business Media Deutschland GmbH, 2020. 91-99.

BibTeX: Download