400 kHz Spectral Domain Optical Coherence Tomography for Corneal Imaging
The cornea is the transparent, outermost layer of the human eye that contributes approximately 70% of the refractive power of the eye in air. It is composed of five major tissue layers: the epithelium, the Bowman’s membrane, the stroma, the Descemet’s membrane, and the endothelium. Corneal diseases such as Keratoconus and Fuchs’ dystrophy can change the morphology of some or all of the corneal layers, which can lead to vision impairment and eventually blindness. For example, Keratoconus causes localizes thinning and thickening of the corneal epithelium, damage to the collagen structure of the corneal stroma (scarring) and alteration of the corneal curvature. All of these changes result in blurred and double vision, and in severe cases can lead to corneal blindness that would require corneal replacement surgery. Fuchs’ dystrophy is a genetic disease that damages the endothelium cell. Since the endothelial cells are responsible for maintaining the fluid level in the stroma, impairment or death of the endothelial cells leads to dehydration or edema of the cornea that results in partial or full corneal blindness. Systemic diseases such as diabetes also affect the physiology and morphology of the cornea. Diabetes affects all the corneal cells and leads to abnormalities such as neuropathy, keratopathy, stromal edema, decrease in endothelial cell density, low tear secretion etc. Although there have been many clinical studies of these diseases, knowledge of the early-stage changes in the corneal morphology at the cellular level remains unclear. Understanding the early stage of disease development with the help of high speed and ultra-high resolution optical coherence tomography (UHR-OCT) corneal imaging can improve the early diagnostics of corneal diseases and well as monitoring the effectiveness of different therapies such as surgical intervention or administration of pharmaceutical drugs. The main objectives of my research project were: a) to upgrade the 34 kHz OCT system with a new camera that offered a 400 kHz data acquisition rate and 8192-pixel linear array sensor, b) test the performance of the 400 kHz OCT system for ex-vivo and in-vivo corneal imaging, and c) develop pre-processing for the interferogram and post-processing algorithms for the images. Implementing a camera with a faster acquisition rate will help to reduce the motion artifact caused by involuntary eye motions. Also, compared to 4500 pixels used in the 34 kHz camera, the new system utilizes approximately 7500 pixels, resulting in a larger scanning range. Although new camera has smaller sensor size (30% smaller), vertical binning is applied to ensure the light signal is all captured. However, due to the faster acquisition rate (~11 times faster), about 10 dB of SNR will suffers from the reduced integration time. Doubling the sample arm power while keep all other conditions the same can boost the SNR by about 3 dB. Therefore, incident power at the sample arm will be raised carefully according to the maximum permissible exposure calculated using the American National Standard for Ophthalmics – Light Hazard Protection for Ophthalmics instruments provided by ANSI. The result from the technical tests shows that the 400 kHz SD-OCT system offers 1 µm axial resolution in biological tissue with an extended scanning range of 2.8 mm (compared to 1.2 mm of the 34 kHz system). It has a lateral resolution of 1.04 μm/pix and can resolve group 7 element 6 of the USAF target with a 20x objective. It can provide 83 dB SNR with 0.95 mW of incident power at a 400 kHz image acquisition rate which should be sufficient to image semi-transparent biological tissues such as the human retina and cornea. However, given the much higher image acquisition rate (> 10x higher), the imaging power can be increased safely to ~ 4 mW, which will increase the system’s SNR to ~ 90 dB. So far, the performance of the 400 kHz OCT system has been tested by imaging plant tissues (cucumber) and ex-vivo pig corneas, due to the cancellation of all in-vivo human and animal studies imposed by COVID-19.
Cite this version of the work
Lin Kun Chen (2021). 400 kHz Spectral Domain Optical Coherence Tomography for Corneal Imaging. UWSpace. http://hdl.handle.net/10012/17821