With the development of Terahertz (THz) sources and detectors, the non-destructive scanning and imaging technique has gain popularity from various fields of applications, from astronomical sciences to product quality inspection, from security and defense applications to pharmaceutical and biomedical applications. Because of its sensitivity to polar molecules, like water, non-ionizing effect and moderate spatial resolution, THz imaging has drawn great attention from medical field, especially in mapping water distribution and migration in physiological tissues.
Given cornea’s low physiologic variation and homogenous composition when compared to other tissue systems in the human body and its easy access with free-space THz beams, corneal tissue water content (CTWC) evaluation with THz waves becomes interesting and popular. As ongoing clinical studies show, a lot of corneal diseases and vision problems are accompanied with abnormal CTWC, appearing as swelling of eyes and blurring of vision, etc. A considerable amount of people all around the world are suffering from different types of corneal diseases. Fuch’s endothelial dystrophy, for example, is affecting around 4% of population who are over 40 years old in the US. As endothelium begins to fail, cornea will get hyper hydrated and impair vision. Late identifications usually requires surgery. Failure of corneal transplant is also believed to be preceded with edema, which means the abnormal CTWC. This makes CTWC mapping an attractive method for doctors to diagnose corneal diseases with. The lack of accurate, non-invasive tool for direct measurement of CTWC opens a great opportunity for THz imaging.
Based on the group’s previous efforts, this thesis aims to develop a proof of concept of rapid scanning THz imaging system for CTWC measurement for study and early evaluation of corneal diseases. In this thesis, it details the modeling of cornea-THz interaction, optics design, system instrumentation, preliminary imaging and system analysis/characterization.
First, a classical electromagnetic model of multi-layer dielectric slabs were adopted for cornea-THz interaction. Observed the reflective signal response based on variations of CTWC and corneal central thickness (CCT) with sources centered at 100 GHz and 525 GHz respectively. Also explored the possibility of more accurate measurement combining magnitude and phase detection.
Second, explored the strengths and weaknesses of imaging with OAPs and found ways to use OAPs correctly. Based on this, the optics setup of a non-contact angular scanning imaging system was designed and fully simulated with ASAP. Much better image quality was achieved with this system compared to the previous system built by the group. Approaches of quasioptics, physical optics and Gaussian beam propagation method were used independently for analysis of the system and had a good agreement with each other.
Based on the optics design, a functioning THz imaging system was implemented afterwards. The electronic schematic of illumination source, detector, preliminary signal processing and data acquisition were explained. Carefully detailed mechanical design, machining and assembling of the system guaranteed the system to run as expected. Custom designed graphic user interface (GUI) enabled operators to control the system and acquire images with ease and efficiency.
System imaging quality and speed limit were characterized with preliminary images. The coupling efficiency of the system were characterized with a brass ball and verified our previous simulation results, proved the system has a homogenous energy coupling efficiency across the effective field of view (FOV). Images with strips-attached brass balls and knife-edge measurement with hemisphere target ball characterized the spatial resolution to be around 1.3 mm with little distortion, agreed well with the predication made via different simulations. Images of cornea phantoms showed the capability of the system to identify abnormal geometry and water migration within the phantoms. The mechanical speed limit of the current system is 8 seconds per image and could be sped up to 1.3 seconds per frame with optimized motion pattern. The electronic setup is limiting the system’s speed to 10 seconds per frame with fine resolution. We believe this optical setup will be able to provide images at video rate with better mechanical and electronic equipment in the future.