3D finite element modeling and signal processing technique of broadband frequency modulated functional near infrared system for human head and detection of traumatic brain injuries

Abstract

The future of non-invasive medical imaging depends on many aspects, such as accurate diagnosis based on high spatial resolution, reliability, cost, and mobility. Imaging using near IR photons reveals accurate assessments of hemodynamic and organ functionality, this can be achieved by accurate extraction of optical absorption and scattering parameters of oxygenated and de-oxygenated hemoglobin. Medical imaging techniques have shown great deal of confidence in improving diagnosis and clinical care of patients suffering from traumatic brain injury (TBI) due to sport, auto accidents, or ordnance blast. Physical injury that leads to temporarily or permanently impaired brain functionality is the main cause of TBI. Due to lack of mobility and high cost of these imaging modalities, non-invasive optical imaging using functional near infrared (fNIR) is being developed by understanding the relation between migrated photons and biological media. Modeling behavior of broadband (30-1000 MHz) frequency modulated near infrared (NIR) photons through a multi-layer human head phantom reveals information for different penetration depth and is of interest to optical bio-imaging of inhomogeneous multilayer tissue. Photon dynamic predictions in human head phantoms using 3D Finite Element Modeling (FEM) are considered fast and accurate scheme for the use of inverse problem solving to predict formation of hematoma or edema in head. The goal of this PhD research was to accurately model the modulated photon behavior when traveling in multi-layer media and inhomogeneous complexity in order to predict early signs of TBI. A FEM simulation of insertion loss (IL) and phase (IP) o are compared to measurements in head phantoms using custom-designed broadband free space optical transmitter (Tx) and receiver (Rx) modules that are developed for photon migration at wavelengths of 680 nm, 795 nm, 850 nm. 2D and 3D FEM based numerical modeling and simulation of IL and IP for a given human head geometry is computed in this thesis based on three layers of phantom each with distinct optical parameter properties to resemble scalp, skull, and cortex. Simulation and experimental results of these phantoms for broadband modulation (30-1000MHZ) are then used for standard error computation at narrowband and broadband frequency modulation between FEM simulated and curve fitted experimental results and confidence in 2D and 3D modeling is achieved. Moreover, the thesis documents a novel signal processing method for early detection of TBI. The modeling and experimental verification of the proposed detection technique is based on FEM and experimental measurements over broadband modulation (30-1000 MHz) for a greater detection accuracy of inhomogeneous multi-layer phantoms representing scalp, skull, and cortex with different sizes of occlusion resembling head with different degrees of hematoma (TBI). Simulation are compared to experimental measurements for identifying minimum spatial resolution based on the proposed signal processing of first and second derivatives of…