The assignee for this patent application is
Reporters obtained the following quote from the background information supplied by the inventors: "Optical Coherence Tomography (OCT) is a technology for performing high-resolution real time optical imaging in situ. In Frequency Domain OCT (FD-OCT), the interferometric signal between reference light and the back-scattered light from a sample point is recorded in the frequency domain rather than the time domain. After a wavelength calibration, a one-dimensional Fourier transform is taken to obtain an A-line spatial distribution of the object scattering potential (A-scan). The spectral information discrimination in FD-OCT can be accomplished by using a dispersive spectrometer in the detection arm in the case of spectral-domain OCT (SD-OCT) or rapidly tuning a swept laser source in the case of swept-source OCT (SS-OCT). Laterally scanning the sample beam over a series of adjacent A-scans synthesizes cross-sectional images, creating a 2-D tomogram commonly called a B-scan. Typically, volumes are acquired by laterally scanning the sample beam over a series of B-scans; however alternative scan patterns, such as a spiral scan patterns, have been suggested to acquire OCT volume data.
"OCT is widely used in ophthalmology to obtain high resolution images of the retina and of the anterior segment of the eye that can be used by doctors to view, diagnose and follow various pathologies in the eye over time. Generally, many closely spaced tomograms around the area of interest are arranged together in a cube/volume format. The histological composition of the retina is a layered arrangement of various tissue types, each having different optical properties relative to the wavelength of the laser used by the OCT device. Consequently, the image representation of the histological structure is a depth (spatially encoded) map of the optical properties of the area being imaged. Visually, the structure of the imaged retina is a curved and layered ordering of intensity. For a given wavelength, the relationship between the intensity and the tissue type it maps to is bound within a statistically defined range for a normal retina, and this specific relationship, along with the specific structure, forms the basis of both the interpretation and processing of the FD-OCT images. Retinal constitution of a typical eye, as imaged by OCT, consists of a layered stack of specific intensities, where each intensity represents a specific tissue type. For example, the nerve fiber layer (NFL) and the retinal pigment epithelium (RPE) appear as a bright layers, and the other layers are of significantly lower intensities.
"A boundary between two different tissue types or layers is an important computational object, and can be automatically estimated as a set of discrete points by intensity based routines. This can subsequently be represented as a smooth surface fitted over the estimated set of points which best discriminate two specific intensity distributions in the image. A boundary, hence, is a smooth surface that separates two specific tissue types, and acts as a surrogate for the layers it separates. The spatial evolution of this boundary surface in space defines the anatomical structure of the retinal layer. Clearly, a boundary as derived above integrates both the intensity and structural information in a single mathematical object.
"Pathologies of the eye often present as structural and intensity modifications of the affected area in the OCT images. Further, because of the functional specificity of the various layers of the retina, different pathologies may affect only a specific subset of the various layers, while sparing the rest of the retinal constitution. This results in a change in a spatial relationship between various retinal layers, and quantifying this change is often a good indicator of an evolving or developed ocular pathology.
"One well established analysis technique, the RPE elevation analysis implemented in the commercially available Cirrus HD-OCT (
"Since there is an implicit integration of structure and intensity during the interpretation of FD-OCT retinal images, automated methods for processing, interpretation and diagnosis of such images can benefit by an explicit extraction of the structure of the eye through the use of differential geometrical tools. The structural information can be used alone, or it can be used to append and enhance what can normally be achieved and understood by intensity alone. Similar types of analysis can also be performed for extra-retinal structures of the eye, for example, the anterior segment structures, including the cornea, crystalline lens and the iris."
In addition to obtaining background information on this patent application, VerticalNews editors also obtained the inventors' summary information for this patent application: "It is an object of the present invention to use geometrical primitives to characterize various regions of 3D OCT image data of the eye. The method relies on intensity based routines to automatically identify regions of interest within a volume of OCT data. In the preferred embodiment of the present invention, representative axial locations of tissue within the eye representing pixels belonging to a specific retinal boundary or a collection of retinal tissue boundaries of interest are identified. With such a selection made for all axial columns in the entire image volume, one gets a set of data points distributed in the 3D volume, and the gestalt of the specific spread of points defines a specific boundary or surface. One skilled in the art may employ other approaches to get a classification of specific point localizers, which can be based on local searches, or volumetric approaches for classification based on intensity distribution in the entire imaged volume.
"The choice of a specific boundary or boundaries for consideration can be determined by the pathology under examination or by input from the system's operator. An alternative embodiment of the invention derives an omnibus analysis of all retinal layers to permit a more exhaustive map of normality or retinal pathology based on geometric properties. A further alternative embodiment could use closed surfaces including the internal or external surface of a lesion or a retinal or choroidal vessel.
"Once a set of points representing the tissue of interest has been obtained, a smooth surface can be interpolated over the points to yield a dense representation of the specific boundary as it varies across the area that has been imaged as illustrated in FIG. 1 for the retinal pigment epithelium (RPE). While only one B-scan is illustrated in the figure, data from multiple B-scans extending across the volume of data would be used to generate the smooth surface. This surface provides a mathematical model of the geometry of the tissue, and can be used to calculate local differential geometric properties of the tissue. As indicated in FIG. 2, the local differential properties can be combined to quantify the surface into regions (parcellations) having specific topographical structures in terms of common differential geometric primitives, i.e. concavities, convexities, bumps, dents etc. Knowledge from clinical practice indicates that many retinal pathologies have a visual appearance detectable in OCT images where specific layers appear deformed and distorted beyond their normal appearance. The invention described herein attempts to link these pathological deformations with differential geometric primitives to which they bear a close resemblance, and hence quantify, classify and localize the affected portion of the imaged retina. The generalized method and its application to a few pathological conditions will be described in detail below.
BRIEF DESCRIPTION OF THE FIGURES
"FIG. 1 illustrates a smooth surface fit to an identified retinal layer according to one aspect of the present invention.
"FIG. 2 illustrates how different geometrical primitives can be used to characterize a retinal surface according to an aspect of the present invention.
"FIG. 3 illustrates the steps of a generalized embodiment of the present invention.
"FIG. 4 is a diagram of a generalized frequency-domain OCT system for use in ophthalmology. [delete 'prior art' from Figure]
"FIG. 5 illustrates how local curvatures can be determined on a surface on a point by point basis.
"FIG. 6 illustrates the sequence of operations required to go from a surface to the geometrical primitives using various combinations of principal curvatures.
"FIG. 7 illustrates an ordering of geometrical primitives based on a particular combination of principal curvatures or shape index according to one aspect of the present invention
"FIG. 8 illustrates the retinal shape of a typical staphyloma case, showing the highly curved shape of the retinal features, including the RPE.
"FIG. 9 illustrates local shape anomalies in the usual shape of the RPE and neighboring layers characteristic of PEDs.
"FIG. 10 shows an OCT B-scan containing a serous PED.
"FIG. 11 illustrates the layers of interest in the anterior segment that would be used to apply the present invention to the analysis of keratoconus."
For more information, see this patent application: SRIVASTAVA, Siddharth; BAGHERINIA, Homayoun. Differential Geometric Metrics Characterizing Optical Coherence Tomography Data. Filed
Keywords for this news article include:
Our reports deliver fact-based news of research and discoveries from around the world. Copyright 2014, NewsRx LLC
Most Popular Stories
- Tesla's Alt-Energy Future Aims for Massive Lithium-Ion Battery Production
- FAA to Appeal Court Decision Allowing Commercial Drone Use
- New Chat App, Yik Yak, Causes Problems for Students
- Rand Paul Tops Presidential Straw Poll at Conservative PAC Conference
- Obama Meets with Ukraine Prime Minister Wednesday
- Gas Prices May Jump from Calif. Emissions Law