The esophagus plays a crucial role in the functioning of the gastrointestinal tract and esophageal disorders are associated with reduced quality of life. Several studies have reported that there is a high worldwide prevalence of esophageal disorders. Esophageal disorders are related to the mechanical properties and function of the esophageal wall. Therefore, to understand the underlying fundamental mechanisms behind various esophageal disorders, it is crucial to map mechanical behavior of the esophageal wall to altered bolus transit and increased intrabolus pressure. This thesis demonstrates how mechanics can be used for better diagnosis of esophageal disorders, improved treatment planning, and identifying the unique mechanical characteristics of various esophageal disorders. To begin with, an in silico model is presented to estimate the optimal parameters of an esophageal surgery called myotomy to reduce the risk of formation of a complication called blown-out myotomy (BOM). Furthermore, two frameworks are described that work with the esophageal diagnostic tests: fluoroscopy and magnetic resonance imaging (MRI), to estimate the mechanical "health" of the esophagus through a set of mechanics-based parameters such as esophageal wall stiffness and active relaxation. These frameworks not only provide valuable information about the underlying wall mechanics behind the measurements of the diagnostic devices but also increase the capabilities of these diagnostic technologies either by adding automation or by increasing the resolution of the images generated by them. Finally, a framework called mechanics-informed variational autoencoder (MI-VAE) is described which generates a mechanics-based parameter space called virtual disease landscape (VDL). Clusters of various esophageal disorders and normal function form within the VDL which help capture similarities and dissimilarities between various diseases. This also helps in identifying mechanical physiomarkers that distinguish the various esophageal disorders. In addition, the VDL helps in estimating the effectiveness of treatments and tracking patient condition over time. Together, these frameworks provide a formal approach by which mechanics can be meaningfully used for diagnosis and treatment planning.

Creator

• Halder, Sourav

DOI

• https://doi.org/10.21985/n2-cv65-ne50

Subject

• Mechanics
• Artificial intelligence
• Biomechanics

Language

• en

Alternate Identifier

• http://dissertations.umi.com/northwestern:16567
• etdadmin_upload_985717

Date created

• 2023-01-01

Resource type

• Dissertation

Rights statement

• In Copyright