TRD1: Functional Imaging

Project Summary
TRD 1
The overarching theme of TRD 1 is to develop powerful new tools for functional imaging: specifically, the devel-
opment of advanced post-processing methods to 1) enhance endogenous contrast in polarization-sensitive
OCT tissue polarimetry, and 2) transform OCT angiography from a pseudo- to a fully-3D technique by dramati-
cally improving its cross-sectional quality and depth resolution. Current post-processing in OCT polarimetry
(PS-OCT) and angiography (OCTA) imparts significant resolution loss—roughly one order of magnitude—gen-
erating contrast with poor spatial resolution compared to the originating OCT data. There is a need for process-
ing techniques capable of preserving spatial resolution to enable a broad range of applications that are
presently outside the reach of OCT technology. This project will develop a probabilistic processing framework
based on the physics of light scattering and OCT image formation that leverages from the similarity of func-
tional signals typically present in biological sample tomograms. This framework will allow for estimation of the
underlying structure and associated contrast—or function—without compromising spatial resolution.
Aim 1 addresses the need for high-resolution cross-sectional polarimetric imaging of the living retina. Proba-
bilistic PS-OCT offers a new pathway to preserve the originating hardware’s resolution, alleviating the need for
increased optical resolution, limited by pupil size in human and animals and accompanied by an impractical re-
duction in the depth of field. This novel capability, combined with PS-OCT hardware equipped with adaptive op-
tics, will enable the determination of polarimetric parameters of individual axonal bundles in the retina. It will
permit PS-OCT to sensitively track retinal axonal degeneration in vivo, thus accelerating the development of
neuroprotective agents for the treatment of multiple sclerosis that rely on rodent models of this disease.
Aim 2 addresses the need to improve the spatial resolution and quality of OCTA in both preclinical and clinical
applications by further extending the probabilistic framework to the OCT signal dynamics. OCTA is also show-
ing significant promise in ophthalmology for the potential use in the early diagnosis and monitoring of diseases
including glaucoma and diabetic retinopathy; preclinical use include imaging animal models to improve under-
standing of tumor biology. However, its poor cross-sectional quality and resolution restricts OCTA to a pseudo-
3D imaging technique, with a depth resolution most commonly four to eight times poorer than in the originating
OCT tomogram, thus limiting angiographic analysis to en face and layer projections in virtually all applications.
Probabilistic OCTA will improve the lacking depth resolution of conventional OCTA, overcoming the insensitivity
to small capillaries and the distortion of true vessel dimensions, which currently undermine caliber and vascu-
lar-network quantitative metrics of great clinical interest in ophthalmology. OCTA with high cross-sectional qual-
ity would amplify its utility in the understanding of the three-dimensional tumor microenvironment, and unlock
the power of volumetric vascular-network metrics in preclinical and clinical applications.

Grant Number: 5P41EB015903-15
NIH Institute/Center: NIH
Principal Investigator: Brett Bouma