Curvature and Color Correction of Ultrawidefield Optical Coherence Tomography - Project Summary/Abstract (30 Lines) While optical coherence tomography has been the predominant diagnostic modality for retinal disease, it still has several shortcomings; commercially available widefield OCT does not show accurate curvature due to fundamental distortion limitations, making monitoring of progression of disease, or in some cases surgical planning, limited. Another problem is the lack of color information from OCT, which relies on single wavelength near-infrared interferometry. As chorioretinal color is an important diagnostic feature of many diseases, or to monitor for progression, best practice in the retina clinic still involves routine pupillary dilation and clinical examination, supported at times by color fundus photography or scanning laser ophthalmoscopy. This K23 grant aims to address both problems, with the goal to allow for high resolution, complete reproduction of the retina with a single imaging modality. We will apply a complete geometric reconstruction of the retina using advanced graphics processing unit (GPU)-accelerated ray tracing techniques. We will also apply the curvature corrected techniques learned from UWF-OCT to perform accurate curvature correction for 2D fundus photos and scanning laser ophthalmoscopy, to enable more accurate geometric measurements for these modalities, improving the quality of clinical trial endpoints and quantitative biomarkers. I have also developed a novel algorithm termed color-OCT, which uses a multiple wavelength light simulation on pure structural OCT data to generate color fundus images. I have shown that it can accurately reflect some disease features, such as the gray-white ROP ridge, the redness of pre-retinal hemorrhage and retinal neovascularization. Given the utility of color in retinal diagnostics, and traditional OCTs unsolved weakness in being unable to detect color, I will seek to validate this new technology on clinical OCT datasets to observe whether it will be valid for common diseases such as age-related macular degeneration, diabetic retinopathy, and optic disc pallor. My primary career goals are to become an independent clinician-scientist exploring computational models and high complexity image processing. My clinical focus is pediatric retina, and I believe many of these modalities will be broadly useful in the rare diseases which afflict this vulnerable population, as well as improve diagnostics in adult vitreoretinal care. Over the past two years of being supported by foundation funding (Knights Templar Eye Foundation, Collins Medical Trust), as well as my mentors funding, I have further developed my computational skills, laying the groundwork for what I would investigate in the K23 grant period. Utilizing these novel software improvements to OCT technology, I hope to improve the quality of retinal diagnostics, and subsequently decrease the dependence on individual clinician examination skills, which may be highly variable. By doing this I hope to level the playing field for access to the highest quality care, as clinical OCT has already done for patients with macular disease and glaucoma.
