**2. Evolution of OCT technology**

Originally, OCT technology was used to localize faults in fiber optic networks in the telecommunications industry [1]. A back light was reflected from a disruption in a fibreoptic cable, and the time-of-flight information was extracted by low-coherence interferometry to enable distance mapping [2]. OCT was introduced in ophthalmology in 1991 as an application of low-coherence interferometry used in axial length

measurement [3]. Other developments that added much to the industry of OCT were semi-conductor, super-luminescent diode SLD (low cost and low maintenance light source), optical heterodyne detection (improving safe detection and interpretation of faint signals) and high-speed line camera technology (incorporating fast repetitive A-scans to give the 2D images "B-scan"). Then, James Fujimoto and his coworkers were the first to represent brightness by employment of the color codes [2].

The first commercially available OCT device was by Zeiss in 1996 after the scanning patterns with reproducible measurements were implemented. In 2000, the second OCT generation was fairly acceptable in ophthalmology practice due to slow speed (100 axial scans/second) and low resolution [2]. Then, the speed was increased in 2002 to 400 axial scans in the third-generation devices for scanning of the anterior segment, retinal layers and optic nerve, and these A-scans became incorporated into a commercial system with an axial resolution of ∼10 μm2 [4]. In 2006, the faster Fourier domain approach using 27,000 A-scans/second was introduced (RTVue, OptoVue, USA).

The adaptive optics AO to correct the ocular monochromatic aberrations was first reported in OCT by Miller et al. in 2003 improving the transverse resolution [5]. However, this AO had a narrow depth of focus that prevented simultaneous visualization of layers at different depths and restricted field of view.

The polarization sensitive was used in OCT in 2001 to measure birefringence of the RNFL of the monkeys [6], and then later on, it was used for birefringence measurement of the retina and anterior segment.

## **3. OCT: basic principles and preclinical application**

OCT is a type of "optical ultrasound" as it relies on time-of-flight information, similar to the ultrasound. It is like an in-vivo optical biopsy analyzing the signals from different locations at different depths [7]. The OCT resolution fills the gap between the ultrasound and conofocal microscopy resolutions.

Light interference is the core principle in OCT imaging based on the optical fiberbased Michelson setup. When the light emits from a low-coherence source, it will split by a coupler into two parts traveling to both arms, namely sample and reference arms. To control different beam parameters (as shape, depth of focus and intensity distribution), the light emitted from the fiber end of either arm is shaped by various optical components (mirrors, lenses, etc.). The backscattered light from both arms will pass through the coupler to be recorded by the detector (**Figure 1**). The difference in the light backscattered from the sample arm will exit when it encounters an interface

#### **Figure 1.**

*Optical fiber-based Michelson setup demonstrating the basic concept of light interference in OCT imaging (figure taken from Popescu et al. [8]).*

between structures of different refractive indices. The corresponding interference pattern is formed between the light propagating in the reference arm traveled a certain optical distance and the light that traveled the same optical distance along the sample arm (including the portion of the distance traveled inside the sample). The depth information of light backscattered from the locations of various structures within the sample can be measured in this way. The detector records the OCT signal during a complete travel of the reference mirror "Depth or A-scan" which is recorded at each beam position. While the B-scan is formed by a consecutive set of A-scans [8].

The 3D imaging allows not only analysis from different locations but also incorporation of A-scans to form the OCT fundus (en face) image [4]. This is called full-field OCT (FF-OCT) with an axial resolution of ~1 μm2 , similar to that of the conventional microscopy [8]. Also, the built-in tracking system allows better sensitivity and higher reproducibility.

In OCT imaging, the axial and transverse resolutions are independent. The axial resolution (depth resolution and coherence gate) is the coherence length of the source which is inversely proportional to its bandwidth. Therefore, broadband optical sources are required to achieve high axial resolution. While the transverse resolution is determined by the minimum spot size of the focused probing beam (inversely proportional to the numerical aperture NA of the focusing lens). This transverse resolution affects the depth of field (a low NA with a greater beam diameter will offer a large depth of field, as in most OCT imaging). The transverse resolutions used by the commercially available OCT systems are between 20 and 25 μm2 [8]. The lateral resolution is considered to be equal to the illumination spot size on the retina (14 μm for Spectralis OCT) [9].
