Retinopathy of prematurity (ROP) is a proliferative retinal vascular disease that affects preterm infants and a major cause of childhood blindness. ROP is diagnosed and staged under indirect ophthalmoscopy at the junction between the vascularized and avascular retina (i.e., periphery) using features such as retinal detachment (RD) and increased vascular dilatation and tortuosity. While surgical and pharmacologic therapies exist, poor structural and visual outcomes (<20/200 vision or blindness) occur in >50% of severe cases because RD, schisis, and vascular abnormalities are often missed on clinical examination. Conventional ophthalmoscopic examination in infants is performed using fundoscopy. However, retinal microvasculature is poorly visualized even when combined with exogenous fluorescein contrast. Optical coherence tomography (OCT) is currently the gold- standard for ophthalmic diagnostic imaging in adults and developments in OCT angiography (OCT-A) have enabled in vivo imaging of retinal vasculature without the need for exogenous contrast. While several research groups have developed handheld OCT/OCTA imaging systems and demonstrated imaging in ROP patients, longitudinal quantitative imaging of retinal vasculature to track ROP progression remains limited by several key challenges: 1) infants cannot fixate, making repeated imaging of regions-of-interest (ROIs) impossible; 2) handheld imaging coupled with infant motion results in significant artifacts and poor OCTA contrast; 3) OCT/OCTA quality is severely degraded by vitreous/anterior chamber haze, which is common in ROP; 4) peripheral retinal vascular changes are important for ROP staging but aiming of OCT/OCTA at these ROIs is difficult; 5) vascular volumes provide clinically relevant features such as dilatation and tortuosity, but OCTA is conventionally assessed using only en face projections. To overcome these limitations, we have developed a combination of hardware and image-processing technologies for handheld OCT/OCTA built around a multimodal spectrally encoded coherence tomography and reflectometry (SECTR) ophthalmic imaging platform. SECTR simultaneously acquires orthogonal en face reflectance and cross-sectional OCT images that uniquely benefits volumetric registration for motion-correction and multi-volumetric mosaicking. We hypothesize that the translation of these technologies into a point-of-care ophthalmic imaging system will allow for robust, reproducible, and quantitative longitudinal tracking of retinal vascular changes, which will improve the diagnostic and staging accuracy of ROP in preterm infants. We aim to develop custom hardware (AIM 1) and image analysis pipelines (AIM 2) optimized for handheld SECTR imaging in infant eyes. These technologies will be validated in longitudinal imaging of structural and functional changes in ROP (AIM 3) and provide quantitative insight on the viability of SECTR imaging to benefit treatment decisions. The proposed device and quantitative analysis pipeline are not limited to ROP cases and can be broadly applied in any area of ophthalmology where a robust point-of- care OCT/OCTA may improve current clinical standard-of-care.

Limited detection sensitivity to features prognostic of retinopathy of prematurity progression by conventional diagnostics is the critical barrier to timely clinical intervention. This program aims to develop robust point-of-care hardware and software technologies to improve the current clinical standard-of-care in preterm infants.