Compared with MRI and optical microscopy, OCT realizes high resolution while the axial resolution and transverse resolution of OCT depend on the different factors, as shown in Figure 2.

Imaging depth plays a crucial role in many OCT applications, particularly in the biomedical field. In the TD-OCT system, the reference mirror position determines the imaging position, making the TD-OCT system able to image at any depth in the ideal case. However, the absorption and scattering of the sample limit the distance of the light path, so the imaging depth is also limited. Without considering the properties of an examined sample, the imaging depth of SD-OCT is determined by the center wavelength of the light source and the spectrometer’s resolution. As shown in Figure 4, the axial resolution of the OCT at 1.06 μm wavelength is better than that at 1.3 μm wavelength. The resolutions of the two OCTs are 11 and 16 μm, respectively. However, at the same imaging depth, OCT at 1.3 μm wavelength has better contrast, and the imaging edge is more obvious because OCT at 1.3 μm wavelength operates at a longer wavelength having higher penetration depth. It also needs to be noted that more noise and interference will be introduced as the depth increases. In theory, the maximum imaging depth, z m a x , can be expressed as z max = λ 2 4 nδλ (4) where δ λ represents the resolution of the spectrometer. This formula proves that a high-resolution spectrometer can improve the imaging depth of SD-OCT.

While considering the properties of a sample, the absorption and scattering make it impossible to reach the theoretical value. Besides, the disordered scattered light will reduce the imaging contrast, system resolution, and imaging depth which causes a decline in image quality. In conclusion, the maximum imaging depth depends on the parameters of the OCT system and the absorption and scattering of the sample.

The sensitivity of the imaging system is also a crucial parameter in biomedical imaging [37]. Since the interferometry that OCT applies measures the distribution of the interference field rather than the intensity of light, the OCT system possesses high sensitivity compared with other optical imaging systems.

Sensitivity, S, is defined as the inverse minimal sample reflectivity in OCT [16], which can be expressed as S =   1 R S , min | SNR = 1 (5)

However, as imaging depth increases, the sensitivity of OCT also rolls off. It can be seen from Figure 4 that as the imaging depth increases, the contrast and sharpness of the images gradually deteriorate. Since the sensitivity of the two OCTs shown in Figure 4 is similar, the difference in imaging performance is mainly due to different wavelengths.

In addition, in SD-OCT, roll-off in the sensitivity is mainly due to the limitation of the number and size of CCD and other components that make up the spectrometer. These limitations cause signal distortion and limit imaging depth. Many attempts have been made to reduce the sensitivity roll-off. For instance, in 2020, a telecentric f-theta imaging lens was designed to decrease the roll-off, and the imaging depth was increased by 13% [38]. After that, David et al. used two reference arms to increase the sensitivity by 14 dB [39].

High imaging speed is a necessary condition for realizing 3D volumetric imaging. Initially, the imaging speed of the first TD-OCT is seconds/a B-scan [1], which is mainly limited by the motion of the reference mirror. Much work has been devoted to improve the delay light path and increase the speed of mechanical motion. Until the emergence of FD-OCT, the imaging speed of OCT became a qualitative leap since spectral information replaced mechanical movement. For FD-OCT with a single beam, the speed of SD-OCT is mainly determined by the imaging time of CCD or CMOS, while the speed of SS-OCT depends on the wavelength-tuning speed of swept sources [16]. In ophthalmology imaging, faster imaging speed can reduce motion artifacts caused by eye movements and thus obtain more explicit images. In Figure 3, the imaging speed of SD-OCT is 175 times that of TD-OCT. The improvement of SD-OCT imaging quality is due to the increase in axial resolution and imaging speed enhancement. The higher A-scan speed makes the edges of different retina layers smoother, and at the same time, reduces movement artifacts caused by eye movement during imaging. The improvement of laser tuning speed, the design of multi-beam path, and the parallel imaging methods have continuously improved the imaging speed of OCT, and the OCT system possessed A-Scan rate as “MHz-OCT” or “multi-MHz-OCT” has been produced [40].

Glaucoma and age-related macular degeneration are the primary diseases that cause blindness [41]. It is crucial to study the vitreoretinal interface comprehensively and further improve our understanding of the pathogenesis and development of retinal diseases. The retinal nerve sensory layer from the outside to the inside is the rods, the cones, the outer limiting membrane layer, the outer nuclear layer, the outer plexiform layer, the inner nuclear layer, the inner plexiform layer, the ganglion cell layer, and the inner limiting membrane [42].

Different eye diseases affect the cells and functional layers in the retina [43]. Hence many diseases can be found from fundus examination. Fundus examinations usually use ophthalmoscopes with limited magnifications, which is insufficient for early observation of small lesions. However, OCT can obtain three-dimensional imaging of the sub-surface of the eyeball structure with the micron spatial resolution and with millisecond time resolution [41, 44], which can observe the pathological alterations of the ocular fundus and provide supplemental information to monitor the progress of the disease accurately and further understanding of the blindness. There is no doubt that ophthalmology is the most widely used application of OCT. OCT has been a standard detection tool in ophthalmology clinics and significantly improves the potential of early diagnosis, monitoring, and precise treatment [33, 45]. The representative OCT image of the mouse retina is shown in Figure 5A, which indicates that OCT can image the layers of the retina clearly, thereby providing corresponding structural information for retinal diseases.

In addition, the extensions of optical coherence tomography angiography (OCTA) also play an important role in visualizing blood flow, as shown in Figure 5B. It is possible to observe the capillaries of the retina and choroid non-invasively, thus acquiring the location, shape, and structure of the diseased blood vessel and evaluating the velocity of blood flow to understand the progress of the disease in depth [46, 47]. The commonly used clinical imaging method of ocular blood circulation is fundus fluorescence angiography. However, this method requires the injection of a fluorescent agent and can only obtain two-dimensional images, which is inferior to OCTA in image resolution, safety, and imaging depth [48, 49]. The angiography results obtained from OCTA have a higher pixel density and fewer motion artifacts and can also measure blood oxygen or provide information about neurological diseases [50, 51]. Besides, OCT can provide more comprehensive imaging and then get the oxygen metabolism of biological tissues combined with multi-spectral polarizing film dispersion while obtaining molecular level biochemical information with Raman spectroscopy [52, 53].

With the development of ultra-broadband laser sources and high-speed detection technology, the ophthalmic OCT system performance has been improved significantly, bringing the potential to 3D noninvasive optical biopsy of the retina in vivo. The morphology in situ obtained by the OCT system is close to the resolution of histopathology [54]. The significant improvement of axial resolution and imaging speed enables three-dimensional high-resolution imaging of the retina and makes it possible to quantify the volume of pathological changes in the retina [31]. Fourier-domain FF-OCT has recently replaced traditional OCT and becomes a powerful tool for ophthalmic imaging due to the highest imaging speed and its capability to avoid motion artifacts and risks in vivo [55]. Moreover, a combination of OCT and hyperspectral imaging allows obtaining the spectral information of oxygen saturation and hemoglobin while acquiring the tissue morphological characteristics [56]. The ability to obtain vascular morphology and functional characteristics is of great significance to clinical applications. Furthermore, the emergence of machine learning makes the automatic processing of images become a reality that can quickly obtain adequate information and make a diagnosis [57, 58].

It is complex progress to convert sound into hearing perception. The sound energy propagates in the cochlea as a traveling wave [59], and the cochlea amplifies low-level sound in a nonlinear active way. The measurement of the progress of sound conduction makes it possible for researchers to have a deep understanding of hearing. However, the vibration inside the inner ear (the cochlea) is still not fully understood. Thus, the following contents mainly introduce the applications of OCT in the inner ear, especially in the cochlea. Moreover, the challenge is to measure nanometer-scale motion in a wide frequency range through the complicated cochlear fine structure [60].

Before the widespread use of OCT, the primary technologies for detecting the cochlea are microscopy, MRI, and CT, but each technique has its drawback. Microscopy, which was initially used to observe the movement in the cochlea, can provide high spatial and temporal resolution. However, the most significant limitation of the microscope is that the imaging depth is insufficient, so it is mainly used to observe the surface and cannot image the cochlea in vivo. Although MRI and CT have great imaging depth and have become standard diagnostic equipment, proving an excellent diagnostic tool for brain, liver, kidney, other solid organs, heart, and large blood vessels, the lack of spatial resolution hinders its application in measuring the cochlear micromechanics. Thus, to study the cochlea deeply, the imaging method requires high-resolution and enough imaging depth. OCT has a micron-level resolution, millimeter-level imaging depth, low cost, and safety, making it an effective method for studying cochlear micromechanics.

Currently, the application of OCT on the cochlea mainly includes three aspects: structural imaging, angiography, and vibration measurement. The first application of OCT on the cochlea was performed in 2007, which acquired images and vibration measurements of the organ of Corti (OoC) of the guinea pig in vivo [61]. Following researches also obtained the cochlear images in mice through this approach [62]. The utility of OCT breaks through the limitation that the laser interferometer can only measure the basal region of the cochlea, which can easily obtain two-dimensional images of each turn of the cochlea and reconstruct the three-dimensional structure images [63, 64]. However, due to the resolution limitation, the internal structure of the organ of Corti cannot be imaged clearly. In recent research, a micro-optical coherence tomography (μOCT) with a high axial resolution of 1 μm obtained from an ultra-broadband supercontinuum laser source was shown to provide subcellular details of the guinea pig cochlea, such as scala tympani, basilar membrane, and other structures or OoC. It proves that the μOCT can distinguish the microstructure of the mammalian cochlea [65], as shown in Figure 6A. Considering the bone of the cochlear capsule, the external auditory canal, and other structures have to be removed during the imaging process, which inevitably causes a certain degree of damage. Therefore, this imaging method is only used to observe the isolated human cochlea [66] and has not yet been applied in living humans. At the same time, it may be used to avoid residual hearing loss during cochlea implant (CI) insertion [67] and may provide real-time 3D images in the future.

OCT angiography (OCTA), a functional extension of general OCT, enable the three-dimensional visualization of the cochlear blood flow by distinguishing the OCT signals of moving red blood cells and static organization without dye, which is significant to the prevention and treatment of many diseases in otology, such as insufficient or abnormal blood supply in the cochlea [68, 69]. Laser Doppler flowmetry (LDF), based on the Doppler principle, is already diagnostic equipment commonly used in modern medicine, while the spatial resolution is limited [70]. Compared with LDF, OCTA possessed a high-resolution is capable of imaging the capillaries of the cochlea.

In addition to the above two applications, OCT can obtain the vibration characteristics at all the cochlea turns and accurately measure the vibration of different structures, such as OHCs, Deiters’ cell regions, tectorial membrane, and the reticular lamina. Compared with a laser doppler vibrometer, OCT can observe the cochlea through the round window membrane, avoiding the placement of the reflective beads. The measurement results of cochlear vibration are shown in Figures 6B,C where (B) is a schematic diagram of Corti, and (C) is the vibration map obtained under 65 dB SPL multi-tone stimulus. We can draw the following conclusions from Figure 6. Firstly, OCT can measure the map of vibration magnitudes; secondly, it was evident that different structures have different amplitudes for the same sound intensity, which implies the sound-evoked motion. The results of other research are as follows. The frequency tuning at the base is sharper than that at the apex [64, 71, 72], and different structures of the OoC have different amplitudes, and phases [73–75], which indicates that the mechanical response of OoC has a transition along the length of the cochlea and the cochlear mechanical response varies with frequency and sound pressure. When measuring the insensitive cochlea, the amplitude increases almost linearly with the intensity of the stimulus, and the different structures of the OoC almost have the same amplitude [76, 77]. However, OCT cannot be highly reliable for the measurement results under very low sound pressure, such as below 20 dB SPL, due to the instrument noise, in which case the signal and the noise threshold are almost at the same order magnitude.

Although OCT is a powerful tool for studying the hearing mechanism, current research is mainly for animals since the cochlea in vivo imaging or vibration measurement will inevitably cause damage. Therefore, further developments will be needed for research or diagnosis of the living human cochlea. In addition, the high resolution and sufficient imaging depth, both limited by the central wavelength, are difficult to meet simultaneously. A recent study that used OCT with center wavelengths at 1,060 nm and 1,300 nm was carried out to explore better wavelengths for cochlear imaging based on contrast and attenuation. It was concluded that 1,060 nm is better for high contrast and enough imaging depth [78]. However, when observing the living cochlea, anatomical conditions and the influence of lymph fluid need to be considered. At this time, a light source with a center wavelength of 1,060 nm may not provide sufficient imaging depth. Moreover, anesthesia also affects the cochlear response, which is confirmed by many current studies. Thus, there is still a long way to go for a thorough understanding of the cochlear response.

Endoscope combined with OCT, another widespread application of OCT, has become an essential tool for in vivo imaging of internal organs. Endoscopic imaging creates the possibility to obtain information from places where are hard to study using traditional methods. In contrast, endoscopic optical coherence tomography (EOCT) can penetrate deeper into the tissue for high-resolution images of the microstructure, such as colon polyps, lung cancer [79]. Another prominent application area of EOCT is biopsy guidance with probe [80]. Thus, EOCT plays a crucial role in detecting, diagnosing, and monitoring organ diseases and provides a guideline for doctors during diagnosis and surgery (77, 78), which can accurately locate the pathological tissue that needs to be detected and distinguish the normal and abnormal physiological structure.

The central wavelength of EOCT used clinically is generally 1,300 nm with an imaging depth between 1 and 2.5 mm, suitable for evaluating the superficial tissue. However, to break the resolution limitation on EOCT and achieve ultrahigh-resolution OCT, the light source centered at 800 nm is better [81, 82]. Furthermore, incorporating the ultrahigh-resolution OCT with a gradient-index (GRIN)-lens micro-endoscope enables visualization at the cellular scale deep in living tissue [83], which is expected to study the development process of specific diseases in vivo and increase clinical applications.

OCT was first introduced into dermatology in 1997 by Welzel et al., who used this technique to image human skin. Skin biopsy and histopathological examination are the gold standards for detecting skin cancer [84, 85]. However, invasive biopsy and a high rate of “false positive” impede the early diagnosis of cancer [86, 87]. Compared with traditional detection methods, OCT with high-resolution can visually the epidermis, dermis-dermal junction (DEJ), dermis, hair follicles, sweat glands, and blood vessels, and other structures and show the morphological characteristics of the pathological structure in the skin [88], which is suitable for imaging the superficial layers of the skin.

Since skin is opaque, the beam will be scattered and absorbed in the propagation process, while the absorption is much smaller than the scattering at the near-infrared spectrum. Thus, commercial OCT used in dermatology always operates in the NIR region [89]. The new OCT system can better visualize the skin structure [90, 91]. It can diagnose different subtypes of skin cancer and accurately determine the tumor boundaries, which helps diagnose disease and clinical treatment detection.