Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Analyzing Deep Learning Techniques in Accelerated Clinical Brain Magnetic Resonance Imaging for Multiple Sclerosis

Mani, Ashika (2021) Analyzing Deep Learning Techniques in Accelerated Clinical Brain Magnetic Resonance Imaging for Multiple Sclerosis. Master's Thesis, University of Pittsburgh. (Unpublished)

[img] PDF
Restricted to University of Pittsburgh users only until 10 May 2023.

Download (766kB) | Request a Copy

Abstract

Background: Magnetic resonance imaging (MRI) scans are routine clinical procedures for monitoring people with multiple sclerosis (MS). Accelerated MRI scan time is motivated by patient discomfort, timely scheduling, and financial burden associated with conventional MRI scans.
Objective: We examined the application of a deep learning (DL) model in restoring the image quality of accelerated clinical brain MRI scans for MS.
Methods: We acquired fast 3D T1w BRAVO and fast 3D T2 FLAIR MRI sequences alongside conventional scans. Using a subset of the scans, we trained the DL model to generate images from fast scans with quality similar to the conventional scans and then applied the model to the remaining scans. We calculated clinically relevant T1w volumetrics (normalized brain volume, normalized thalamic volume, normalized gray matter volume, and normalized white matter volume) for all scans. We performed paired t-tests for conventional, fast, and fast with DL for these volumetrics, and fit repeated measures linear mixed-effects models to test for differences in correlations between volumetrics and clinically relevant patient-reported outcomes. We performed equivalence tests to compare fast scans with DL and conventional scans to examine equivalence in image quality as well as equivalence in association with patient-reported clinical outcomes.
Results: We found statistically significant differences between conventional scans and fast scans with DL for all T1w volumetrics. There was no difference in the extent to which the key volumetrics and clinical outcomes are correlated between fast scans with DL and conventional scans, but there was not sufficient evidence to prove that the correlations were equivalent.
Conclusion: There is currently no evidence to support that fast scans with DL produce images of equivalent quality to conventional scans. However, fast scans with DL have the potential to inform clinically relevant outcomes in MS.
Public health significance: Limited research has been done regarding the application of deep learning models to improve the image quality of accelerated scans in clinical brain MRIs for people with multiple sclerosis. The results of this analysis can inform practitioners as to how to further incorporate and improve on MRIs utilizing deep learning.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Mani, Ashikaasm134@pitt.eduasm134
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Thesis AdvisorYouk, Adaayouk@pitt.eduayouk
Committee MemberJeanine, Buchanichjeanine@pitt.edujeanine
Committee MemberCarlson, Jennajnc35@pitt.edujnc35
Committee MemberZongqi, Xiazxia1@pitt.eduzxia1
Date: 10 May 2021
Date Type: Publication
Defense Date: 26 April 2021
Approval Date: 10 May 2021
Submission Date: 26 April 2021
Access Restriction: 2 year -- Restrict access to University of Pittsburgh for a period of 2 years.
Number of Pages: 56
Institution: University of Pittsburgh
Schools and Programs: Graduate School of Public Health > Biostatistics
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: magnetic resonance imaging, deep learning, linear mixed-effects models, equivalence tests
Date Deposited: 10 May 2021 22:49
Last Modified: 10 May 2021 22:49
URI: http://d-scholarship.pitt.edu/id/eprint/40805

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item