Loading...
Detection and classification of cerebral vascular injury via microwave sensing and AI
Jensen, Reilly Shawn
Jensen, Reilly Shawn
Citations
Altmetric:
Files
Loading...
t25043_Jensen.pdf
Adobe PDF, 4.34 MB
Authors
Other Names
Location
Time Period
Advisors
Original Date
Digitization Date
Issue Date
2025-07
Type
Thesis
Genre
Keywords
Subjects (LCSH)
Electronic dissertations
Citation
Abstract
This thesis evaluates the capability of a monostatic RFR in detecting physiological changes which accommodate cerebral hemorrhage. The expected result is that the data collected from the RFR can be used in the training of machine learning models which can then accurately predict the flow rate of hemorrhage. The second chapter of this thesis is dedicated to a literature review covering the following: The physiology of hemorrhagic stroke, alternative monitoring methods and their underlying theories and applications, current obstacles and future works. The third chapter summarizes the methods and results of the dataset generation and artificial intelligence model development and performance.
This study used an ex vivo porcine model. Radio frequency sweeps were conducted using a vector network analyzer. The machine learning models used were: linear regression, logistic regression, random forest, k-nearest neighbors. Several convolutional neural network models were explored. Hyperparameter searches were conducted, and the best performing models were evaluated. It was concluded that linear regression and logistic regression failed to predict flow rate based on resonant frequency shifts. Both the k-nearest neighbors and random forest models performed well, achieving 81.1% and 74.2% accuracy, respectively. Finally, the results for the convolutional neural networks were inconclusive due to insufficient dataset size.
Table of Contents
Description
Thesis (M.S.)-- Wichita State University, College of Engineering, Dept. of Biomedical Engineering
Publisher
Wichita State University
Journal
Book Title
Series
Digital Collection
Finding Aid URL
Use and Reproduction
© Copyright 2025 by Reilly Jensen
All Rights Reserve