Advanced Machine Learning Techniques for Analysis of 4D Brain Images in Ischemic Stroke Diagnosis and Assessment

• Author / Creator

  Soltan Pour, Mohsen

• Acute Ischemic Stroke (AIS), a devastating cerebrovascular disorder, is one of the
  leading causes of disability and mortality worldwide. It occurs when the blood supply
  to a part of the brain is interrupted, resulting in the deprivation of oxygen and nutrients, leading to neuronal damage and functional impairment. With its widespread
  occurrence and significant impact on individuals and society, ischemic stroke substantially burdens healthcare systems. The World Health Organization (WHO) estimates that stroke affects approximately 13.7 million individuals worldwide annually, of which nearly 85% are attributed to ischemic strokes. Moreover, stroke-related deaths account for about 11% of global deaths, making it a critical public health concern.
  Timely and accurate diagnosis and a comprehensive assessment of ischemic stroke are crucial for guiding appropriate treatment strategies and improving patient outcomes. Despite their efficacy, conventional diagnostic methods often depend on subjective interpretation and may not fully capture the dynamic nature of stroke progression. Therefore, there is a growing need for advanced machine learning techniques to aid in the analysis of 4D dynamic brain images, enabling precise diagnosis and assessment of AIS. Computed Tomography Perfusion (CTP) imaging is the most common technique in assessing ischemic stroke. It provides information about cerebral hemodynamics, tissue perfusion, and blood volume. Utilizing the swift acquisition of sequential CT images immediately after the administration of a contrast agent, CTP serves as a valuable technique for accurately identifying at-risk (penumbra) and already damaged (core) viable tissue in an AIS patient’s brain. Quantitative measurements derived from CTP, such as Cerebral Blood Flow (CBF) and Mean Transit Time (MTT), aid in estimating AIS outcomes and guiding treatment decisions.
  This thesis aims to explore and develop novel machine-learning methods for analyzing 4D brain images to enhance AIS diagnosis and assessment, ultimately improving patient care and outcomes. The main objective is to develop an automatic AIS baseline\follow-up lesion segmentation\prediction. By utilizing advanced machine learning techniques and analyzing medical imaging data, accurate identification and assessment of ischemic lesions can be achieved. This information guides effective treatment decisions, determines intervention timing, and predicts outcomes, while also tailoring treatment plans and optimizing resource allocation. Additionally, it aids in clinical trials by identifying suitable patients for targeted interventions, thus enhancing stroke management and benefiting both patients and healthcare systems.

• Subjects / Keywords

  • Acute Ischemic Stroke (AIS)
  • Deep Learning
  • Machine learning
  • medical imaging
  • segmentation

• Graduation date

  Fall 2023

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/r3-emp6-gf09

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.