OCT utilizes the principles of interferometry and coherent light to produce images of both diagnostic and regulatory quality. However, the realm of OCT did not start with medical imaging elegance. The first commercial use of OCT was not in medicine but in the world of art history and old master painters to determine if multiple paintings were layered on a single canvas. Due to the expense of proper canvases, many Old World artists were inclined to paint over previous paintings, entombing the original work. On many occasions, the artwork concealed beneath the surface of the visible image was much more valuable and more important than the surface facade. In a similar way, what lays beneath the surface of the retina and optic nerve are often more important than what is seen with the ophthalmoscope.

Before we explain how OCT has achieved its current sophistication, we will review the fundamental technology. When a wavelength of light encounters the boundary of two different media, the light may be reflected, refracted, and/or absorbed. The first step for OCT is the acquisition of the reflected light by a capture device.

Next, the algorithm of interferometry, an electromagnetic principle, is used. “Interference” can be either constructive or destructive for imaging purposes depending upon the relative phases of the waves. OCT employs the interaction of waves coherent with each other to form images by sectioning or “tomography,” a physical application used in CT, MRI, and ultrasonography.

One of the limiting technological issues for the first OCT instruments was the number of A scans that could be performed in a finite time, thereby limiting image resolution (Fig. 13.3). Improvement in this engineering domain has increased the number of A scans and shortened the capture time for this biomechanical variable.

Another similarity with MRI, specifically orbital imaging, demanded correction for normal, continuous, physiological movements of the eyes in the resting stage, termed microsaccades. Until MRI software conquered this obstacle, orbital MRI scans were blurred and substandard due to movement artifacts. By using a variety of techniques to insure image stabilization and registration, SD-OCT devices have reduced this source of artifact (Fig. 13.4) and hence further improved image quality and decreased the coefficient of variability for measuring the dimensions of the retina.

Because OCT requires light to be precisely focused upon the retina, imaging is often limited by the dimensions of the eye. Somewhat smaller than normal eyes, hyperopia (far sightedness) greater than 6 diopters, and somewhat larger eyes, myopia (near sightedness) greater than 6 diopters, may induce errors in image resolution and the measurement algorithms of the peripapillary retinal nerve fiber layer thickness and the macular thickness.

When TD-OCT was introduced into the commercial market, the technology was patent protected. As SD-OCT emerged, patent protection was no longer possible because Fourier analysis was required to improve image quality. Fourier analysis is considered to be common scientific knowledge, not subject to patent enforcement. SD-OCT facilitated competition among device manufacturers in the commercial market, which has produced improved technology and helped to control price. The hardware and software platforms from various manufacturers differ fundamentally in methodology of image processing as well as the definition of the posterior border of the retina. Does the retina end at the inner border of the retinal pigment epithelium (RPE), the outer border of the RPE, or Bruch’s membrane? The answer to that question depends upon which device a clinician or clinical trial selects (Tables 13.1 and 13.2). The most important conclusion from the different definitions of the borders of the retina becomes the inability to transfer and merge qualitative data from one OCT platform to another for determining baseline measurements as well as changes over time during a clinical trial. Measuring percentage change over time may be a way to approach this conundrum but that is somewhat flawed since the basic dimensions are different, and this approach has not been accepted by regulatory authorities. Tables 13.1 and 13.2 further illustrate this problem, which becomes very significant for international trials where use of a single device is rarely possible unless supplied by the sponsor. Ultimately, regulators may declare which platforms and measurements algorithms are acceptable for specific trials.