1. Introduction

Optical coherence tomography (OCT) imaging technology has developed rapidly since it was first introduced in 1991[1]. To obtain video rate imaging, the rapid scanning optical delay line (RSOD) was developed[2, 3]. OCT equipped with RSOD technology enabled many in vivo biomedical imaging applications, which require real-time image acquisition[4–7]. Recently, Fourier domain OCT, which makes use of spectral interferometry has further increased imaging speed [8–10]. Development of very broad-bandwidth light sources, including femtosecond lasers and supercontinuum generated in optical fiber, have resulted in axial resolution as low as 0.5 μm[11–13]. To adapt OCT technology to various clinical applications, a variety of scanners have been developed[14]. Previous OCT scanner development has focused on design of the scanners for specific applications (e.g. catheter probes for endoscopic imaging), not on optimization of the optical design.

In the presence of optical aberrations, a beam scanned off of the optical axis can be seriously distorted. In the context of an OCT scanner, this distortion can result in deterioration of the transverse resolution as well as inefficient recoupling of the backscattered signal light. For a bench-top OCT scanner, it is desirable to achieve a transverse scan range of at least several millimeters. This wide range results in significant beam distortion unless care is taken to minimize aberrations using optimized optical design.

The investigation of OCT image quality variation becomes more important in the development of three dimensional OCT (3D OCT) with large scale lateral scans. One way to achieve 3D OCT imaging is to scan the beam in the two transverse dimensions using two orthogonal scanning mirrors that are physically separated. This kind of configuration of 3D OCT scanning is well established and commonly used in many applications [15–18]. In a telecentric optial system, all of the chief rays on the image side are parallel to the local optical axis and therefore perpendicular to the image plane [19]. Telecentric optics generally feature little aberration, which leads to several advantages for an OCT scanner, including constant magnification, a constant spot size off axis, and a flat image plane[20]. Therefore, an OCT scanner with telecentric optics can achieve consistent image quality over a wide lateral scan range. In order to achieve a telecentric scan, the scanning mirror must be placed at the front focus of the optics that follow. This design is not difficult to achieve for a 2D OCT imaging system because only one scanning mirror is needed. However, if a scanner for 3D OCT makes use of an x-y scan head, which has two separated orthogonal scanning mirrors, it is obviously impossible to simultaneously align both at the front focus of the optics.

This problem can be addressed in several ways[21]. For example, a straightforward solution is to insert relay optics between the x and y scanning mirrors[22]. A well designed relay lens system images the x-scanning mirror to the y-scanning mirror, so that the y-scan and the image of the x-scan are both at the front focus of the objective. However, this design makes the system more complex and less compact. Another approach is to use a single mirror with two tilting axes[23]. These devices are recently commercially available, but are currently limited to small tilt angles and slow scan rates compared to galvanometer-mounted scan mirrors. One method that has been suggested to reduce telecentric error is to mount one of the x-y scanning mirrors off of the rotating axis[24]. However, the movement of the mirror placed off of the axis includes both shifting and rotating, which changes the optical path length. Therefore, this configuration is not optimum for an OCT scanner.

The approach we present in this report is to make use of an x-y galvanometric scan head, which is commercially optimized and widely used in OCT. We present a design, that we call quasi-telecentric, which approximately achieves the advantages of a true telecentric design. In this approach, the two separated mirrors lie on either side of the front focus of the optics that follow. The mirrors lie at small and approximately equal distances from the focus, compared to the front focal length of the optics. This optimized quasi-telecentric design is able to reduce the off-axis aberration to an acceptable level and achieve a uniform image quality through the entire lateral scan range. Furthermore, this OCT scanner includes a view port that accommodates a microscope to precisely guide and document OCT imaging of small samples.

2. Design strategy

The goal of the reported design was to develop an OCT scanner with high lateral resolution, long working distance (the distance between the final optical component and the sample), and large lateral scanning range. In addition, we attempted to achieve a uniform spot size and image quality over the entire lateral scan range. We did not address the distortion of the geometrical location of the beam because it is a compensable parameter by post imaging processing[25]. Another design goal was to provide a view-port for a microscope to visualize the sample simultaneously with OCT imaging. The primary purpose of this scanner is for use with an OCT system intended for bench-top studies of biomedical samples under the guidance of a microscope.

The light source of the OCT system used with this scanner has a coherence length of 10μm (full width at half maximum) in air which determines the axial resolution of the OCT image. In a three dimensional OCT system, the lateral resolution should ideally match the axial resolution. Therefore we required that the spot size at the focus should be not larger than 10μm (full width at half maximum) or 17μm (at 1/e²) diameter. The lateral scanning range should be wide enough to give the investigators a sufficient field of view to find the target of interest, examine it, and to compare it with the surrounding environment. However, design difficulty is increased if lateral scanning range is too large. We selected a design goal of 4–5 mm lateral scan range, representing sufficient field of view for bench-top samples of interest to us. The working distance should be long enough for the users to conveniently place and handle the sample under the scanner. However, a longer working distance requires a larger beam size before the objective optics to maintain a small spot size at the focus, which is more susceptible to aberration. We selected a design goal of 15–20 mm working distance, which would adequately accommodate bench-top samples of interest to us. The spot size and quality, lateral scanning range and working distance are the major design criteria that were balanced in optimizing this design. To maximize uniformity of spot quality across the lateral scan range, we selected two quantitative parameters to represent spot quality: spot size variation (ratio of the absolute difference between on- and off-axis spot diameter to the mean of on- and off-axis spot diameters), and spot ellipticity (ratio of the absolute difference between major and minor axes to the mean of major and minor axes). As our design goal, we chose a tolerance of less than 5% for both of these parameters. To indicate the degree of quasi-telecentricity, we defined a quasi-telecentricity parameter, QTP, as the ratio of the half separation between the two scanning mirrors to the front focal length of the optics that follow.

 

Fig. 1. Layout of quasi-telecentric OCT scanner optical design with view port and microscope, modeled by Zemax. AL, aspheric lens; X-Y Galvs, x-y galvanometric scan head; LP_1 and LP_2, achromatic lens pairs; M_1 and M_2, folding mirrors; OB_1 and OB_2, identical objectives; DM, dichroic mirror; IS, 1:1 image of sample (view port).

Download Full Size | PDF

In order to provide a view-port for a microscope, our design makes use of relay optics to image the scan head to a dichroic mirror at the front focus of the objective optics. In order to achieve a quasi-telecentric scanner (where the displacement of the scanning mirrors from the front focus of the optics that follow is small compared to the front focal length of the optics that follow) that makes use of a compact scan head, we use beam-expanding relay optics.

The optical design is shown in Fig. 1. The illuminating light delivered via optical fiber is collimated by an aspheric lens AL into a diameter of 2.2mm and then deflected by a commercially available, small and compact x-y galvanometric scan head. The x-y scan head (Cambridge Technology, Cambridge, MA) used in our design provides high scanning speed and has a distance (d) of 5.4 mm between x and y mirrors which leads to a small deviation from telecentric optics. The relay optics consists of two pairs of achromatic lenses, LP-1 and LP-2, which magnify the OCT beam by a factor of two. In order to minimize spherical aberration in a large angle scan, LP-1 and LP-2 are pairs of identical achromatic lenses face to face. The focal lengths of LP-1 and LP-2 are f₁=62mm and f₂=124 mm respectively and they are separated by f₁+f₂. Folding mirrors M1 and M2 are used to make the scanner compact. A dichroic mirror DM with a double pass reflectivity of 98% for 1310nm wavelength at 45 degrees incident angle is employed to deflect infrared light and transmit visible light with a transmission of 80%. It is located at both the back focus f₂ of LP-2, where the image plane of galvanometer is, and the front focus f₃= 20mm of the objective OB-1. It deflects the OCT beam through the objective lens OB-1 which then focuses the beam onto the sample. OB-1 is a combination of singlet and achromatic lenses. It is a simplified method of minimizing spherical aberration similar to LP_1 or LP_2. All optics used in the design are commercially available. The OCT signal reflected or scattered from the sample placed at the back focus f₄ of OB-1 is collected back through the same path to the OCT detector, while visible light passes through OB_1, DM and OB_2 to the microscope and CCD camera. OB-2 is identical to OB-1, so that they constitute a finite conjugate system to provide a 1:1 image of the sample at conjugate plane IS and allows the microscope to indirectly image the sample at IS. This design does not make use of a visible aiming beam to identify the location of the OCT scan in the microscope image. Instead, the location of the OCT scanning is calibrated by two small pinholes and then marked on the screen by software (a “virtual aiming beam”). This design allows investigators to make use of a microscope without additional adaptation.

 

Fig. 2. (a)–(c) Cross section profiles of the spots simulated by Zemax;.(d)–(f) Spot profiles corresponding to (a)–(c) measured by beam analyzer. Mesh grid and Zemax windows are 20×20 μm. Center means the beam goes through the center of the optics (on axis), while x and y mean tilting the x and y mirrors, respectively, to translate the beam in either x or y direction by 2.2mm.

Download Full Size | PDF

3. Spot analysis and characterization

Figures 2(a)–(c) shows the spot profiles on the flat image plane simulated by Zemax (company name) in physical optics propagation (POP) mode. Figure 2(b) is at the center of the field of view on the image plane, in other words it is the on-axis spot profile. Figure 2(a) and (c) are offset from the center by 2.2 mm in the y- and x-directions respectively. The size of the profile windows is 20×20 μm² for all three profiles. To quantify the diameter of slightly elliptical spots, we here report the average of the spot width in the x and y directions. The 1/e² diameter of the simulated central spot was 14.9 μm, while the diameter of the 2.2mm y-offset spot was 15.4 μm, and the diameter of the 2.2mm x-offset spot was 15.3 μm This achieves our design goal of not exceeding 17 μm spot diameter. The maximum fractional spot size deviation is 2.3% over the full lateral scan range of 4.4 mm, which is better than our tolerance criterion of 5%. The average ellipticity of x- and y-scans is 3.2%, which also meets our tolerance criterion of 5%.

We measured the actual spot profiles of the scanner as shown in Figure 2 (d)–(f) using a beam analyzer placed at the image plane, defined as the plane perpendicular to the optical axis at the focus of the on-axis beam, which is 16.5 mm away from the last surface of the objective lens. The measurements (d)–(f) were at the three different positions corresponding to the simulations (a)–(c). The mesh grid used to scale the dimension of the image is 20×20 μm². Fig. 2(e) shows the spot profile of the beam going through the system’s optical axis. The measured 1/e² diameter of the central beam was 14.8 μm, while the diameters of the 2.2mm y-offset beam and 2.2mm x-offset beam were 16.1 μm. This corresponds well to our simulated spot profiles. The spot size remained below our design goal of 17 μm over the full scan range, and the maximum fractional spot size deviation of 4.4% meets our tolerance criterion. The average measured ellipticity of x and y scans in Fig. 2 is 1%, which is better than our tolerance criterion of 5%.

Comparing Figure 2 (a)–(c) with (d)–(f), we note that except for the central beam, the spot size measured by the beam analyzer is slightly larger than the one simulated by Zemax. One possible explanation is that we assumed the core diameter of the SMF-28 optical fiber to be 9 μm, but the diameter of the actual fiber may vary by +/-0.4 μm. Another possible explanation is that the beam analyzer might not be precisely aligned with the image plane. The off-axis spot profiles are slightly elliptical, which is a typical problem for a system with slight spherical aberration. The further the spots are from the optical axis, the more elliptical they become. Another important property shown in Figure 2 is the purity of the spot profile. There are no side lobes, which could decrease the image contrast.

 

Fig. 3. Simulations showing variation of the relative off-axis spot profile with QTP. Diamond and square represent spot size variation and ellipticity respectively

Download Full Size | PDF

In order to investigate the effects of varying the degree of quasi-telecentricity, we varied the distance between the two scanning mirrors in our Zemax model system while keeping the front focal length of LP-1 constant and the focus of LP-1 at the middle distance between the two mirrors (see Fig. 3). The quasi-telecentricity parameter value of zero in Fig. 3 represents a telecentric system. We varied the distance d up to 10mm, which varies the quasi-telecentricity parameter up to approximately 8%. In Fig.3, we plotted the variation of the two spot parameters at the maximum of the lateral scan range as a function of the quasi-telecentricity parameter: the spot size variation, and spot ellipticity. The average values for the x- and y-spots are plotted.

The results shown in Fig. 3 indicate that the relative spot size increases slowly until the QTP reaches approximately 5% then increases more rapidly. The ellipticity generally decreases slowly with QTP. We observe an oscillation in ellipticity with QTP, but have no explanation for this observation at this time. The arrow in Figure 3 represents the QTP of our implemented system, 4.4%. At this value, as reported above, both spot size variation and ellipticity are well below our tolerance criteria.

4. OCT images

Figures 4 and 5 display real-time OCT images of biological samples acquired using the optimized scanner. The real-time OCT engine used here is similar to one previously described[3, 26]. The light source has a bandwidth of 70nm (full width at half maximum) centered at 1310nm. 5mW of optical power was incident on the sample and the images were acquired at 8 frames per second with 500 axial lines (A-scans) per image and a 4 frame rolling average for speckle noise reduction. These examples illustrate the high OCT image quality achievable with the optically optimized scanning system.

 

Fig. 4. OCT images of in vivo human skin recorded using the quasi-telecentric scanner. Image dimensions: vertical 2.8mm, horizontal 3.9mm. Images recorded at 8 frames per second with four-frame rolling average, and displayed without additional processing. (a)—Nail fold; (b)—Thick skin (palm of hand); (c)—Thick skin (finger pad); (d)—Thin skin (back of hand). Note uniform image quality across the lateral field of view. Note fine structures that are not always resolved in OCT images of skin acquired without high quality scanning optics, such as resolution of stratum corneum from living epidermis and epidermis from dermis in thin skin. BV—Blood vessels; DM—Dermis; ED—Epidermis; NP—Nail plate; PNF—Proximal nail fold; SC—Stratum corneum; SD—Sweat duct. All tissue identification is tentative, made by comparison with published histology, in the absence of directly corresponding histology (biopsies of imaged sites were not taken).

Download Full Size | PDF

Figure 4 shows in vivo OCT images of four types of human skin with image dimensions of 2.8 mm axially and 3.9 mm laterally. It is noteworthy that the image quality is maintained over the full lateral range. Figure 4(a) shows a cross-section through the nail fold. The fine structures and small blood vessels in the proximal nail fold at the upper left of Fig. 4(a) are imaged with high resolution by use of this scanner. Differentiation of the components of the nail unit may be important for nail surgery[27]. Figures 4(b) and 4(c) show the structure of human thick skin on the palm of the hand and the finger pad, respectively. Boundaries between the stratum corneum, living epidermis and dermis are distinguished, and sweat ducts are visible as they coil through the stratum corneum. Figure 4(d) shows the structure of human thin skin. Compared to thick skin, the thin skin has much thinner epidermal layers and more dermal structure is visualized. The boundaries between stratum corneum, living epidermis, and dermis are visible in thin skin, as in thick skin. In the dermis, small blood vessels are apparent, especially in 4(c) and 4(d). To resolve the boundary between the epidermis and the dermis may have clinical significance for examination of skin cancers using OCT technology.

 

Fig. 5. OCT images of embryonic chick heart on (A)day 5 (Stage 28) and (B) day 6 (Stage 30). OFT--Outflow tract; tr—trabecular network; epi – epicardium; AVC – atrioventricular cushion; ivs -- interventricular septum; RV—Right ventricle; LV—Left ventricle; CM—Compact myocardium.

Download Full Size | PDF

Figure 5 shows OCT images of the embryonic chicken heart (A-stage 28 and B-stage 30). These images also demonstrate the high resolution and contrast enabled by the optimized scanner. The thin epicardial layer (epi) covering the compact myocardium (CM) up to the border between the myocardial and mesenchymal regions of the outflow tract is clearly visible with a distinct boundary. Clearly resolving the trabecular network in the ventricles is crucial for accurately determining cardiac parameters such as ejection fraction, stroke volume, and wall thickness. Other noteworthy features include the atrioventricular cushion (AVC) in Fig. 5(A) and the inter-ventricular septum (IVS) in Fig. 5(B). These features form during cardiac septation. Mechanisms for malformations of these structures are not clearly understood and can be studied with our system. These high resolution images, acquired in real-time, will enable accurate anatomical and physiological measurements on these very small, developing hearts. This scanner has been important to our on-going project developing gated OCT technology for developmental cardiology research[28].

5. Conclusion

We have described a two-dimensional OCT scanner using orthogonal scanning mirrors that is approximately telecentric. We demonstrate the importance of optimizing the optical design in creating such devices. We provide a design criterion for the reported scanner, that the ratio of the half distance between the two scanning mirrors to the front focus of the following optics (QTP) is less than 0.05. This design criterion succeeded in satisfying our design goals of 5% tolerance in both the spot size variation and spot ellipticity over the entire lateral scanning range of 4.4mm in both x- and y-dimensions. The view port enables a microscope to be coupled to the scanner for monitoring and documenting OCT imaging procedures. The design strategy has been validated by high quality OCT images in biological samples of interest to our research group.