Microwave Imaging of Neck for Early Detection of Alzheimer’s Disease: Numerical Feasibility Study

Alzheimer’s disease (AD) is the most common form of dementia, affecting millions of people worldwide and posing increasing challenges to healthcare systems. Early, noninvasive diagnosis is a paramount medical need. This thesis contributes to exploring the feasibility of Microwave imaging (MWI) as a possible solution. When exposed to microwaves operating at about 1 GHz, the scattering response of healthy and AD-affected tissues alters due to dielectric contrast arising from physiological changes in the Cerebrospinal fluid (CSF), a tissue that supports and protects the central nervous system. By conducting a preliminary numerical analysis, the performance of an MWI setup in various pathological scenarios that gradually worsen has been estimated. Simulations were performed on a digital twin made of a 3-D anthropomorphic model composed of tissue-mimicking layers, with pathology modeled via small permittivity changes in the CSF region, and an array of probes that act like both microwave emitters and receivers, designed to surround the model’s neck. Using a Finite Element Method (FEM) solver, the electromagnetic fields in both healthy and pathological scenarios have been estimated. Then, to reconstruct dielectric contrast volume distributions as a function of the FEM results, the Truncated Singular Value Decomposition (TSVD) algorithm was selected. The preliminary results demonstrate promising capabilities for discriminating pathological conditions in the early stages of AD, suggesting that the proposed technique could complement the existing diagnostic methods. Further research is needed to refine the digital twin and validate findings with experimental data.