2.1.

High-Speed Spectral Domain OCT System

Figure 1 illustrates the schematic of the high-speed SD-OCT system used in this study, which is similar to one described previously.²³ Briefly, light from a $56\text{-}\mathrm{nm}$ -bandwidth, low-coherence broadband infrared superluminescent diode light source $(1300\pm 28\mathrm{nm})$ was split into two paths in a 10:90 fiber-based Michelson interferometer. One beam was coupled onto a stationary reference mirror, and the second was focused and scanned using a pair of galvo mirrors. The light emerging at the output of the interferometer was sent to a custom-built high-speed spectrometer. The spectrometer consisted of a transmission grating $(1175\text{lines}∕\mathrm{mm})$ , a camera lens with a focal length of $100\mathrm{mm}$ , and a 1024-element line scan infrared InGaAs detector with a $25\text{-}\mu \mathrm{m}\text{pixel}$ pitch size. The spectral resolution of the designed spectrometer was around $0.141\mathrm{mm}$ , which provided a total depth range of $3.3\mathrm{mm}$ in air ( $2.45\mathrm{mm}$ in biological tissue by assuming the refractive index of the sample was $\sim 1.35$ ).

Fig. 1

Experimental setup of high-speed SD-OCT system: $\mathrm{SLD}=\text{superluminescent}$ diode, $\mathrm{OC}=\text{optical}$ circulator, $\mathrm{FC}=10:90$ fiber coupler, $\mathrm{PC}=\text{polarization}$ controller, $\mathrm{L}1–\mathrm{L}5=\text{lenses}$ , $\mathrm{RM}=\text{reference}$ mirror, $\text{LASER}=\text{pilot}$ laser for beam guiding, $\mathrm{DG}=\text{diffraction}$ grating, $\mathrm{CCD}=\text{line}$ -scan camera, $\mathrm{COMP}=\text{computer}$ .

One main requirement of the high-speed SD-OCT application is utilization of the camera’s full acquisition capability. In our system, the integrating time was set at $17\mu \mathrm{s}$ , and the maximum line rate of the camera was $47\mathrm{KHz}$ . CameraLink^™ and a high-speed frame-grabber board (PCI 1428, National Instruments, USA) were used to transfer the raw fringe data from the camera to the host computer. A custom-designed, high-speed software was developed under the Lab VIEW 2009 platform for real-time data acquisition, processing, display, and archiving. The axial resolution of the system was $13\mu \mathrm{m}$ in air ( $\sim 9.6\mu \mathrm{m}$ in tissue). We used an objective lens with a $50\text{-}\mathrm{mm}$ focal length to deliver the probe beam of $4\text{-}\mathrm{mm}$ diameter onto the sample, providing a $\sim 16\text{-}\mu \mathrm{m}$ transverse resolution (in tissue) and a $\sim 2\text{-}\mathrm{mm}$ depth of focus. The estimated SNR of the system was around $90\mathrm{dB}$ with a light power on the sample arm at $3\mathrm{mW}$ .

2.2.

Image Acquisition and Reconstruction

For in vivo imaging, our imaging protocol provided 3-D cochlear imaging at 280 B-scan frames per second (fps). Each B-scan encompassed a $1.5\mathrm{mm}\times 2.45\mathrm{mm}$ $\left(x\text{-}z\right)$ slice of cochlea. A series of 125 cross-sectional B-scan images were obtained from the medial to lateral direction $\left(1.25\mathrm{mm}\right)$ , and only $\sim 0.45\sec$ was required to capture a 3-D volume of $1.5\mathrm{mm}\times 1.25\mathrm{mm}\times 2.45\mathrm{mm}$ ( $128\times 125\times 512$ voxels with a voxel size of $\sim 12\times 10\times 10\mu {\mathrm{m}}^3$ ) that covered the cochlea.

Before applying the traditional SD-OCT algorithms to reconstruct the OCT images, all the spectral interferograms in each B scan along the $x$ direction were ensemble-averaged at each wavelength to obtain a reference spectrum, which was then subtracted from each A-scan. This operation effectively removed/minimized the autocorrelation, self-cross-correlation, and camera noise artifacts presented in the final OCT images,²⁴ and considerably improved the image quality. The subtracted spectral interferograms were then converted into the equal frequency space by use of the spline interpolation method, because the spectral interferogram data matrices captured by the CCD camera were functions of wavelength; however, the Fourier transform relationship was between time (distance) and frequency (wavenumber). After scaling the depth range by the refractive index of 1.35 for biological tissue, the reconstructed 2-D OCT images were stacked sequentially to form a 3-D volume with a physical size of $1.5\mathrm{mm}\times 1.25\mathrm{mm}\times 2.45\mathrm{mm}$ $(X\times Y\times Z)$ .

2.3.

Animal Model and Surgical Technique

Since the penetration depth of our OCT system to access the cochlea was limited by the presence of the thick, bony mastoid bulla on the cochlear surface, surgical access to the cochlea was required to visualize the cochlear microstructure. Healthy C57 BL/6 mice weighing around $20\text{to}30\mathrm{gm}$ ( $3\text{months}$ old) were used in this study. The experimental protocol was in compliance with the federal guidelines for care and handling of small rodents and approved by the Institutional Animal Care and Use Committee (IACUC) of Oregon Health and Science University (OHSU).

Animals were initially anesthetized by a single subcutaneous (s.c.) injection of ketamine $(75\mathrm{mg}∕\mathrm{kg})$ and xylazine $(10\mathrm{mg}∕\mathrm{kg})$ . The mouse was then placed on a heating pad and maintained at $37\pm 1°\mathrm{C}$ . Prior to surgically exposing the cochlea, the mouse head was immobilized and positioned onto an imaging platform to minimize the translation of movements due to breathing, which is essential for in vivo imaging. The imaging platform was based on a standard mouse stereotaxic instrument.

Under sterile surgical conditions, a ventral incision approach with a longitudinal, paramedian incision was used to access and remove the bone of the tympanic bulla to permit direct access to the inner ear with minimal blood loss or mortality.²⁵ This approach provided a more direct view of the cochlear turns and round window niche area compared to the retroauricular approach. After incision, the left submandibular gland and posterior belly of the left digastric muscle were removed by cauterization to reveal a well-defined sternocleidomastoid muscle and facial nerve extending anteriorly above the bulla. The bony bulla was exposed, and the dorsal region of the bulla was surgically removed using a micro scissor to provide a clear view of the intact cochlea and the stapedial artery (SA) with its medial margin laying over the edge of the round window niche and coursing anterior-superiorly toward the oval window. After surgery, the anesthetized mouse with exposed and intact cochlea (Fig. 2 ) was moved under the beam emanating straight down from the SD-OCT imaging probe for image scanning. During imaging, local anesthesia with isoflurane ( $0.2\mathrm{L}∕\min$ ${\mathrm{O}}_2$ , $0.8\mathrm{L}∕\min$ air) was applied, and the breathing interval of the animal was controlled by adjusting the concentration of isoflurane in the vaporizing chamber of the anesthetic station.

Fig. 2

Surgical view of the left cochlea in mouse. (a) Ventral surgical window through the bulla exposes the middle and basal turns of the cochlea (within the rectangular box). White scale $\text{bar}=1\mathrm{mm}$ . (b) Optically magnified view of the exposed cochlea with otic capsule left intact. The dotted rectangle shows the approximate region of the OCT scan. White scale $\text{bar}=200\mu \mathrm{m}$ .

3.1.

Results

In vivo OCT images of a portion of the apical, middle, and basal turns of the mouse cochlea were obtained through a surgically prepared opening through the bone of the bulla. Figure 3 shows the in vivo 2-D OCT cross-sectional image of the cochlea imaged through the intact osseous otic capsule, and Fig. 3 shows a standard cochlear cross-sectional image, that was imaged using orthogonal-plane fluorescence optical sectioning microscopy (OPFOS) described in other publications.^{26, 27} Figure 4 shows an enlarged view of one of the turns shown in Fig. 3. The bony labyrinth of the cochlea has a spiral shape something like the space inside a snail shell. As shown in Fig. 3, the cochlea consists of relatively large fluid-filled spaces referred to as the scala vestibuli (SV), scala media (SM), and scala tympani (ST). Figure 3 also demonstrates that the OCT imaging method is able to detect a wide variety of tissue types ranging from acellular gelatinous material (the tectorial membrane) to dense bone (the otic capsule) to the epithelial organ of Corti, which contains the remarkably delicate hair cells.

Fig. 3

(a) In vivo 2-D cross-sectional OCT image of apical turns of the spiral-shaped cochlea $(1.5\mathrm{mm}\times 2.45\mathrm{mm})$ . Compare this image with the first frame of Video 1. White $\text{bar}=250\mu \mathrm{m}$ . (b) Optical cross-sectional image using OPFOS.²⁶ $\mathrm{O}=\text{otic}$ capsule, $\mathrm{M}=\text{modiolus}$ , $\mathrm{RM}=\text{Reissner}$ ’s membrane, $\mathrm{ST}=\text{scala}$ tympani, $\mathrm{SM}=\text{scala}$ media, $\mathrm{OC}=\text{organ}$ of Corti, and $\mathrm{SV}=\text{scala}$ vestibuli. White $\text{bar}=250\mu \mathrm{m}$ .

Fig. 4

(a) Enlarged in vivo 2-D cross-sectional OCT image of a single apical turn of the cochlea. Both the tectorial membrane (TM) overlying the organ of Corti (OC) and the basilar membrane (BM) are visible in this image. (b) Optical sectioning of a single apical turn of the cochlea using OPFOS²⁷ (reprinted with permission of Elsevier). $\mathrm{O}=\text{otic}$ capsule, $\mathrm{M}=\text{modiolus}$ , $\mathrm{RM}=\text{Reissner}$ ’s membrane, $\mathrm{BM}=\text{basilar}$ membrane, $\mathrm{ST}=\text{scala}$ tympani, $\mathrm{SM}=\text{scala}$ media, $\mathrm{OC}=\text{organ}$ of Corti, and $\mathrm{SV}=\text{scala}$ vestibuli.

Video 1

Flythrough of the 2-D cross-sectional images of the cochlea from the middle to apical turns (QuickTime, $3910.36\mathrm{KB}$ ). .

For a better view of the cochlear structures, the 2-D cross-sectional OCT images from the apical to middle turns are shown in Video 1 . The cochlea is oriented at a slight angle relative to the imaging beam such that the two cross-sectional halves of the apical turn are seen at the upper right-hand corner at the beginning of the video clip. As the video clip progresses through the cochlea, the right-hand cross-section of the turn ends at the most apical end of the cochlea, called the helicotrema, and the left-hand cross section continues into the middle turn. Figure 5 (Video 2 ) shows a 3-D volumetric projection image of the cochlear microanatomy obtained by SD-OCT. The dashed line in Fig. 5 indicates the first B-scan. The location of the SA is indicated with the black lines, and the direction of the C-scan is shown with the dotted arrow. To reveal the internal cochlear turns more clearly, the image transparency was increased with 3-D reconstruction software (Amira). As can be seen in both the 3-D OCT image of the cochlea (Fig. 5) and the cross-sectional OPFOS image [Fig. 3], the diameter of each turn becomes progressively smaller from the base to the apex. The cochlear structures exist in a complex, helical geometry spiraling around a central pillar of spongy bone with nerve fibers, the modiolus (M). In comparison with the standard in vitro OPFOS image, the OCT cross-sectional image presented here demonstrates a similar microanatomy, both in morphology and scale. For example, the thickness of the organ of Corti (OC) measures approximately $80\mu \mathrm{m}$ at its greatest dimension in both the OCT image and the OPFOS image. In addition, the cross-sectional fly-through movie (Video 1) rendered from the 2-D OCT stack images clearly reveals the spiral canal of the cochlea, which is divided lengthwise into three passages (SV, SM, and ST) separated by Reissner’s membrane and the basilar membrane (BM). The SV and the ST are connected to each other at the helicotrema [Figs. 5, 6, 6 ].

Fig. 5

3-D volumetric projection image of the cochlear microanatomy (Video 2). Dashed line indicates the location of the first B scan. The location of the stapedial artery (SA) is indicated with the back lines, and the direction of the C scan is shown with the dotted arrow.

Fig. 6

3-D volume rendering of mouse cochlea, which is segmented and displayed in four different orientations to provide a detailed view of the cochlea. (a) 3-D volumetric image of the entire cochlea imaged. $\mathrm{O}=\text{optic}$ capsule, $\mathrm{M}=\text{modiolus}$ . (b) 3-D reconstructed image of the cochlea in the sagittal plane provides an en face view of Reissner’s membrane and a longitudinal cross-section of the scala vestibuli (SV) and basilar membrane (BM). (c) 3-D reconstruction of the first 50 B frames from the lateral-to-medial direction. (d) 3-D reconstruction of first 25 B frames from the lateral-to-medial direction. Cochlear structures, such as the otic capsule (O), Reissner’s membrane (RM), basilar membrane (BM), modiolus (M), scala tympani (ST), scala media (SM), organ of Corti (OC), and scala vestibuli SV are clearly visualized in (c) and (d).

Video 2

3-D volumetric reconstruction of the cross-sectional OCT images of the apical and middle turns of the cochlea (QuickTime, $185\mathrm{KB}$ ). .

Figure 4 shows the enlarged, detailed 2-D cross-sectional image of the cochlea imaged with OCT [Fig. 4] in comparison with an OPFOS image [Fig. 4] illustrating the cochlear microanatomy. Figure 6 shows several views of 3-D volume-rendering based on segmentation of OCT data in the sagittal and coronal planes. From the sagittally segmented, 3-D–rendered, and resliced OCT image of the cochlear structure, OCT volume-rendering clearly reveals the cochlear duct (SM) located between Reissner’s membrane and the basilar membrane, which can be seen in both Figs. 5 and 6. Figure 6 also shows the ability of OCT to visualize the organ of Corti in vivo, which resides within the cochlear duct on the basilar membrane. The tectorial membrane is adherent to the organ of Corti in the scala media space. The organ of Corti is a highly organized structure that is responsible for the transduction of mechanical vibrations into neural impulses. The organ of Corti is visible in the 3-D OCT reconstruction [Fig. 6]; however, the detail of individual hair cells within the organ of Corti is beyond the resolution of our present system.

3.2.

Discussion

In the present study, the feasibility of high-speed SD-OCT for clear visualization of the inner ear microanatomical features was successfully demonstrated. A tool that provides in vivo high-resolution imaging to reveal anatomic and physiological information of the inner ear in a relatively noninvasive and nondestructive manner might shed light on fundamental issues concerning hearing function and loss, enable clinically significant diagnoses, and guide future surgical efforts to restore hearing function by micromanipulation. The results of the current study of in vivo imaging of cochlear morphology with high-speed SD-OCT suggest that this imaging modality deserves exploration for such applications.

As an alternative in vivo imaging modality for viewing intracochlear structures, SD-OCT has several potential advantages over other cochlear imaging modalities such as CT, MRI, and HR-US. First, SD-OCT is relatively inexpensive and portable. Second, it has the potential to provide real-time images and can be integrated into surgical guiding tools, such as the endoscope and laryngoscope, which might be very useful for inner-ear intervention in both human subjects and animals. Other current in vivo imaging techniques, such as CT and MRI, do not have the sufficient spatial resolution to visualize the microanatomical structures of the cochlea, such as Ressiner’s membrane, the basilar membrane, and the organ of Corti. The ability to visualize Ressiner’s membrane and the basilar membrane in vivo has great importance in both basic scientific research and clinical studies and could be used to understand a wide range of inner ear disorders. For example, endolymphatic hydrops is the result of an enlargement of the endolymphatic fluid space with ballooning of Reissner’s membrane and displacement of the basilar membrane. At present, there are no imaging modalities other than OCT that can provide such high-resolution in vivo imaging of these cochlear microstructures. The ability of OCT to visualize inner-ear morphologies, such as described above, could change our current understanding of them. However, our current system resolution $(\sim 10\mu \mathrm{m})$ is still not high enough to image individual cellular morphologies within the cochlea. Ultrahigh-resolution OCT (UH-OCT) imaging that uses state-of-the-art broad-bandwidth optical sources, such as femtosecond or supercontinuum lasers, in conjunction with a high numerical aperture (NA) microscopic objective can provide sufficient cellular-level resolution to image the hair cells of the organ of Corti.

Another potential application of OCT is studying the mechanics of sound-wave propagation through the ribbon-shaped BM and how it transduces movement to the cellular and acellular components of the organ of Corti, which are attached to the basilar membrane. In vivo imaging also plays a crucial role in cochlear implantation. OCT is feasible for checking the electrode placement in a cochlear implantation. In clinically difficult cases, such as fibrosis in the basal turn or altered anatomy, such as cochlear otosclerosis, it is very relevant to visualize the basilar membrane and the relatively fine position of the cochlear implant to accurately place the electrode.