High-Speed, Low-Cost, Image Remapping Spectral Domain Full-Field Optical Coherence Tomography for Retinal Imaging

点击翻译按钮获取中文摘要

Optical Coherence Tomography (OCT) has profoundly impacted diagnostic imaging of the human retina, enabling accurate and early diagnosis of retinal disease. The relatively high cost of current clinical instruments limits its use in cost sensitive environments, e.g. screening in the optometrist's or general practitioner's office, underserved or rural US populations and developing countries. Motion of the patient and/or interferometer strongly effects phase stability and image quality, impacting the common approaches which scan a focused beam across the retinal surface. Collecting the entire field at once via an approach such as Full-Field Optical Coherence Tomography (FF-OCT), would obviate this issue and provide high spatial phase stability. Here we propose a novel solution that not only overcomes these issues (poor phase stability and coherent artifact) but promises to do so while reducing costs and increasing imaging speed. We will take advantage of rapid development in two key technologies, additive manufacturing for optical components and extremely high megapixel CMOS arrays. Building on concepts from our prior work, we will develop an FF- OCT system by mapping the 3-dimensions (x,y, and k) onto a 2-D CMOS array within an imaging spectrometer. This will in turn allow us to utilize inexpensive spatially incoherent light sources which do not produce coherence artifacts, but heretofore were limited to use with time-domain FF-OCT systems. This will be accomplished via 3 Specific Aims. Aim 1, Develop a 3-D printed structure for image remapping in FF-OCT: We will deploy a micro- optical design consisting of arrays of 3-D printed single mode fibers that map x,y in the image plane, to spaced columns. The space between the columns will enable dispersion in k within the 2-D spectrometer developed in Aim 2. This structure will enable volumetric OCT imaging with a single camera exposure by mapping x, y, and k onto a 2-D array. Aim 2, Develop an imaging spectrometer and spectral domain FF-OCT system around a high megapixel (Mp) CMOS sensor: We will design and develop an imaging spectrometer that will be compatible with the structure(s) from aim 1, integrating the structures as they become available. Aim 3, System validation and human retinal imaging: After validation on tissue phantoms, pilot human data from a spectrum of retinal diseases will be acquired and compared with standard-of-care clinical imaging systems. We expect the completed system to have an FOV of 5 mm diameter, 15.5 µm lateral sampling, 7.5 µm axial resolution and be able to collect a volume image in as little as 200 µs at 48 volumes per second.