Optical tomography uses the backscattering of light to provide cross-sectional images of ocular structures. Visible (or near-visible) light works well for imaging the eye because many important structures are optically transparent (cornea, aqueous humor, lens, vitreous humor, and retina – see Figure U.3-1).
Figure U.3-1 Schematic representation of the human eye.
To provide analogy to ultrasound imaging, the terms A-scan and B-scan are used to describe optical tomography images. In this setting, an A-scan is the image acquired by passing a single beam of light through the structure of interest. An A-scan image represents the optical reflectivity of the imaged tissue along the path of that beam – a one-dimensional view through the structure. A B-scan is then created from a collection of adjacent A-scan images – a two dimensional image. It is also possible to combine multiple B-scans into a 3-dimensional image of the tissue.
When using optical tomography in the eye it is desirable to have information about the anatomic and physiologic state of the eye. Measurements like the patient’s refractive error and axial eye length are frequently important for calculating magnification or minification of images. The accommodative state and application of pupil dilating medications are important when imaging the anterior segment of the eye as they each cause shifts in the relative positions of ocular structures. The use of dilating medications is also relevant when imaging posterior segment structures because a small pupil can account for poor image quality.
Ophthalmic tomography may be used to plan placement of a phakic intraocular lens (IOL). A phakic IOL is a synthetic lens placed in the anterior segment of the eye in someone who still has their natural crystalline lens (i.e. they are “phakic”). This procedure is done to correct the patient’s refractive error, typically a high degree of myopia (near-sightedness). The exam will typically be performed on both eyes, and each eye may be examined in a relaxed and accommodated state. Refractive information for each eye is required to interpret the tomographic study.
A study consists of one or more B-scans (see Figure U.3-2) and one or more instances of refractive state information. There may be a reference image of the eye associated with each B-scan that shows the position of the scan on the eye.
The anterior chamber angle is defined by the angle between the iris and cornea where they meet the sclera. This anatomic feature is important in people with narrow angles. Since the drainage of aqueous humor occurs in the angle, a significantly narrow angle can impede outflow and result in increased intraocular pressure. Chronically elevated intraocular pressures can result in glaucoma. Ophthalmic tomography represents one way of assessing the anterior chamber angle.
B-scans are obtained of the anterior segment including the cornea and iris. Scans may be taken at multiple angles in each eye (see Figure U.3-2). A reference image may be acquired at the time of each B-scan(s). Accommodative and refractive state information are also important for interpretation of the resulting tomographic information.
Figure U.3-2 Tomography of the anterior segment showing a cross section through the cornea.
Note in the Figure the ability to characterize the narrow angle between the iris and peripheral cornea.
As a transparent structure located at the front of the eye, the cornea is ideally suited to optical tomography. There are multiple disease states including glaucoma and corneal edema where the thickness of the cornea is relevant and tomography can provide this information using one or more B-scans taken at different angles relative to an axis through the center of the cornea.
Tomography is also useful for defining the curvature of the cornea. Accurate measurements of the anterior and posterior curvatures are important in diseases like keratoconus (where the cornea “bulges” abnormally) and in the correction of refractive error via surgery or contact lenses. Measurements of corneal curvature can be derived from multiple B-scans taken at different angles through the center of the cornea.
In both cases, a photograph of the imaged structure may be associated with each B-scan image.
The Retinal Nerve Fiber Layer (RNFL) is made up of the axons of the ganglion cells of the retina. These axons exit the eye as the optic nerve carrying visual signals to the brain. RNFL thinning is a sign of glaucoma and other optic nerve diseases.
An ophthalmic tomography study contains one or more circular scans, perhaps at varying distances from the optic nerve. Each circular scan can be “unfolded” and treated as a B-scan used to assess the thickness of the nerve fiber layer (see Figure U.3-3). A fundus image that shows the scan location on the retina may be associated with each B-scan. To detect a loss of retinal nerve fiber cells the exam might be repeated one or multiple times over some period of time. The change in thickness of the nerve fiber tissue or a trend (serial plot of thickness data) might be used to support the diagnosis.
Figure U.3-3 Example tomogram of the retinal nerve fiber layer with a corresponding fundus image.
In the Figure, the pseudo-colored image on the left shows the various layers of the retina in cross section with the nerve fiber layer between the two white lines. The location of the scan is indicated by the bright circle in the photograph on the right.
The macula is located roughly in the center of the retina, temporal to the optic nerve. It is a small and highly sensitive part of the retina responsible for detailed central vision. Many common ophthalmic diseases affect the macula, frequently impacting the thickness of different layers in the macula. A series of scans through the macula can be used to assess those layers (see Figure U.3-4).
A study may contain a series of B-scans. A fundus image showing the scan location(s) on the retina may be associated with one or more B-scans. In the Figure, the corresponding fundus photograph is in the upper left.
Figure U.3-4 Example of a macular scan showing a series of B-scans collected at six different angles
Some color retinal imaging studies are done to determine vascular caliber of retinal vessels which can vary throughout the cardiac cycle. Images are captured while connected to an ECG machine or a cardiac pulse monitor allowing image acquisition to be synchronized to the cardiac cycle.
Angiography is a procedure which requires a dye to be injected into the patient for the purpose of enhancing the imaging of vascular structures in the eye. A standard step in this procedure is imaging the eye at specified intervals to detect the pooling of small amounts of dye and/or blood in the retina. For a doctor or technician to properly interpret angiography images it is important to know how much time had elapsed between the dye being injected in the patient (time 0) and the image frame being taken. It is known that such dyes can have an affect on OPT tomographic images as well (and it may be possible to use such dyes to enhance vascular structure in the OPT images), therefore time synchronization will be applied to the creation of the OPT images as well as any associated OP images
The angiographic acquisition is instantiated as a multiframe OPT Image. The variable time increments between frames of the image are captured in the Frame Time Vector of the OPT Multi-frame Module. For multiple sets of images, e.g. sets of retinal scan images, the Slice Location Vector will be used in addition to the Frame Time Vector. For 5 sets of 6 scans there will be 30 frames in the multi-frame image. The first 6 values in the Frame Time Vector will give the time from injection to the first set of scans, the second 6 will contain the time interval for the second set of 6 scans, and so on, for a total of 5 time intervals.
Another example of an angiographic study with related sets of images is a sequence of SLO/OCT/"ICG filtered" image triples (or SLO/OCT image pairs) that are time-stamped relative to a user-defined event. This user-defined event usually corresponds to the inject time of ICG (indocyanine green) into the patients blood stream. The resultant images form an angiography study where the patient’s blood flow can be observed with the "ICG filtered" images and can be correlated with the pathologies observed in the SLO and OCT images which are spatially related to the ICG image with a pixel-to-pixel correspondence on the X-Y plane.
The prognosis of some pathologies can be aided by a 3D visualization of the affected areas of the eye. For example, in certain cases the density of cystic formations or the amount of drusen present can be hard to ascertain from a series of unrelated two-dimensional longitudinal images of the eye. However, some OCT machines are capable of taking a sequence of spatially related two-dimensional images in a suitably short period of time. These images can either be oriented longitudinally (perpendicular to the retina) or transversally (near-parallel to the retina). Once such a sequence has been captured, it then becomes possible for the examined volume of data to be reconstructed for an interactive 3D inspection by a user of the system (see Figure U.3-5). It is also possible for measurements, including volumes, to be calculated based on the 3D data set.
A reference image is often combined with the OCT data to provide a means of registering the 3D OCT data-set with a location on the surface of the retina (see Figures U.3-6 and U.3-7).
Figure U.3-5 Example 3D reconstruction
Figure U.3-6 Longitudinal OCT Image with Reference Image (inset)
Figure U.3-7 Superimposition of Longitudinal Image on Reference Image
While the majority of ophthalmic tomography imaging consists of sets of longitudinal images (also known as B scans or line scans), transverse images (also known as coronal or “en face” images) can also provide useful information in determining the full extent of the volume affected by pathology.
Longitudinal images are oriented in a manner that is perpendicular to the structure being examined, while transverse images are oriented in an “en face” or near parallel fashion through the structure being examined.
Transverse images can be obtained from a directly as a single scan (as shown in Figures U.3-8 and U.3-9) or they can also be reconstructed from a 3D dataset (as shown in Figures U.3-10 and U.3-11). A sequence of transverse images can also be combined to form a 3D dataset.
Figure U.3-8 Transverse OCT Image
Figure U.3-9 Correlation between a Transverse OCT Image and a Reference Image Obtained Simultaneously
Figures U.3-8 through U.3-10 are all images of the same pathology in the same eye, but the two different orientations provide complementary information about the size and shape of the pathology being examined. For example, when examining macular holes, determining the amount of surrounding cystic formation is important aid in the following treatment. Determining the extent of such cystic formation is much more easily ascertained using transverse images rather than longitudinal images. Transverse images are also very useful in locating micro-pathologies such as covered macular holes, which may be overlooked using conventional longitudinal imaging.
Figure U.3-9 Correspondence between Reconstructed Transverse and Longitudinal OCT Images
Figure U.3-10 Reconstructed Transverse and Side Longitudinal Images
In Figure U3.9, the blue green and pink lines show the correspondence of the three images. In Figure U3.10, the Transverse image is highlighted in yellow.