Characterizing Cerebral Degeneration in Amyotrophic Lateral Sclerosis with Texture Analysis

• Author / Creator

  Ishaque, Abdullah

• Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease that is hallmarked by unrelenting and progressive wasting and paralysis of voluntary muscles. The classic pathology in ALS features neuronal loss, gliosis, and abnormal protein deposition in the upper and lower motor neurons (UMN and LMN). Most patients diagnosed with ALS die within five years of symptom onset due to respiratory failure. Studies have used magnetic resonance imaging (MRI) techniques to capture the neurodegenerative processes in the brain in vivo and have consistently shown abnormalities in the motor cortices, corticospinal tract, and frontotemporal regions. However, the longitudinal course of cerebral degeneration remains poorly understood. Additionally, clinical trials are hampered by a lack of feasible objective biomarkers that can monitor the disease and provide a means to mitigate the detrimental effects of the heterogeneity of disease in ALS.

  In this dissertation, the overall objectives were to study cerebral degeneration in ALS using routinely acquired MR images and texture analysis. Texture analysis is an image processing technique that quantifies variations and relationships between voxel intensities in an image. Secondly, the longitudinal changes in gray and white matter structures in ALS were investigated to study the progressive course of cerebral degeneration in the disease.

  To accomplish these objectives, the dissertation was divided into three experiments. First, 3-dimensional (3D) texture analysis was applied to T1-weighted MR images of ALS patients and controls. Baseline group comparisons identified abnormalities in texture of T1-weighted images in the motor cortex, insula, frontal lobe, basal ganglia, parahippocampal regions, and corticospinal tract in ALS patients. Furthermore, patients with survival of less than 20 months had greater involvement of frontotemporal regions than patients with longer lengths of survival. In the second experiment, the discriminatory accuracy of these texture features from the corticospinal tract was evaluated in two independent cohorts. Additionally, the association between texture of the corticospinal tract from T1-weighted images and its diffusion metrics from diffusion tensor imaging were investigated. Binary logistic regression models using texture features of the corticospinal tract were able to discriminate ALS patients from controls. Several texture features correlated with diffusion metrics and these features were also significantly different along the length of the corticospinal tract, particularly in the regions of the corona radiata and internal capsule, between ALS patients and controls. These experiments highlight the ability of texture analysis to quantify gray and white matter degeneration using a single T1-weighted image sequence. In the last experiment, the longitudinal progression of cerebral degeneration was investigated. Here, progressive texture abnormalities in the internal capsule in ALS patients over four- and eight-month intervals in the absence of clinical UMN decline were found. Clinical UMN dysfunction also correlated specifically with texture of the corticospinal tract. Longitudinal gray matter progressive changes were characterized by a frontotemporal spatial spread, instead of worsening degeneration within the motor structures. Longitudinal findings differed between fast and slow progressing ALS subgroups; the slow progressing ALS subgroup had greater progressive texture change in the internal capsule than the fast progressing ALS subgroup. On the other hand, the fast progressing ALS subgroup had greater progressive texture changes in the precentral gyrus.

  In summary, this dissertation has revealed important insights into the longitudinal gray and white matter degeneration in the brain in ALS and their relationship to their clinical phenotype. Particularly, progressive degeneration along the corticospinal tract occurs over a short period of time in the absence of clinical UMN decline. Over the same period, the longitudinal course of gray matter pathology is characterized by a frontotemporal spatial spread instead of progressive degeneration within the motor structures. Texture analysis of T1-weighted images successfully recapitulated the known in vivo spatial pathological cerebral characteristics of ALS with abnormalities in the motor cortex, frontotemporal regions, and the corticospinal tract. Texture-based abnormalities of the corticospinal tract correlated with clinical UMN dysfunction, highlighting the importance of this structure as a suitable marker for UMN disease. Therefore, texture analysis of T1-weighted images has emerged as a potential objective biomarker due to its feasibility in clinical application. Based on the results, it can serve and fulfil the roles of multiple types of biomarkers including diagnostic and prognostic markers.

• Subjects / Keywords

  • neuroimaging
  • neurology
  • neurodegeneration
  • brain

• Graduation date

  Fall 2020

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/r3-65xc-5r84

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.