Scanning laser ophthalmoscope for small animals

Abstract

A compact confocal scanning laser ophthalmoscope comprises an MEMS scanning mirror with a compact scanning head and is suitable for small animal subjects. The scanning head is connected to the light source, the detector and the control electronics through flexible optical and electrical umbilical members and includes a zoom lens to provide a wide field of view and high resolution imaging of the retina using multiple imaging modes.

Description

Technical Field

[0001] The present invention relates to a confocal scanning laser ophthalmoscope for examining the fundus of small animals (i.e., the inner posterior surface of the eye). Background Art

[0002] Fundus imaging of small animals is increasingly widely used in basic research, drug development, and pre-clinical research for the treatment of ophthalmic diseases. Typical subjects include mice, other rodents, and zebrafish. These animals are used to study visual physiology, as models for the development of human eye diseases, and for the development of treatments for various eye and nervous system diseases.

[0003] Existing instruments include traditional fundus cameras. These instruments are well known and are described, for example, by Sato in US 3,016,000 and more recently by Massie in US 7,993,000 specifically for small animals. The problem with using such direct imaging instruments is the need to simultaneously illuminate the entire field of view on the retina. Light scattering in the lens and vitreous of the subject's eye results in an undesirable loss of contrast.

[0004] Pomerantzeff and Webb described a scanning ophthalmoscope in US 4,213,678 that alleviates these problems by scanning a laser spot focused on the retina such that only a small portion of the field of view is illuminated at any given time. Webb, Hughes, and Delori (1987) described further improvements where they inserted a pinhole spatial filter in the detection path that is confocal with the small area of the retina illuminated by the scanning laser and suppresses light scattered from other surfaces. Blaha and Gaida described another example of such a confocal scanning instrument in US 5,071,246.

[0005] Scanning laser ophthalmoscopes have been used for fundus imaging of small animals. Unfortunately, clinical instruments designed for human subjects are bulky and not well suited for use with smaller subjects. Research instruments such as those described by Zhang et al. (2015) are also inconvenient to use. Therefore, there is a need for an instrument whose ergonomics are suitable for studying small animal subjects.

[0006] The present invention combines the small size and ease of use associated with a small fundus camera, improved optical contrast, high lateral and axial optical resolution, and a wide range of imaging modes that may be achieved by a confocal scanning mechanism.

[0007] A significant feature of a confocal scanning laser ophthalmoscope (cSLO) is the scanning mechanism, which deflects the incident laser beam as well as the scattered or fluorescent signals returning from the subject. Ideally, the incident and returning beams are substantially coaxial and are tilted in unison about a point within the eye (near the pupil center or the anterior nodal point of the eye). This is typically achieved by performing a two-dimensional raster scan using one or more rotating, tilting, or vibrating scanning mirrors. An optical relay is set up such that the tilt axis of the mirror is at the conjugate focus with respect to the beam tilt axis within the subject's eye.

[0008] The detected signal is maximized when the effective diameter of the beam returning to the detector is constrained only by the diameter of the subject's pupil. As discussed by Webb et al. (1987), in the absence of vignetting, for a collimated beam, the product of the beam area and the solid angle swept by the scan is an optical invariant of the system between the pupil and the scanning mirror. The beam width and deflection range at the subject's pupil determine the minimum size and deflection capabilities required of the scanning mirror. In terms of the characteristics of the scanning mirror, the product of the beam area at the scanning mirror and the sine of the orthogonal mirror deflection angle is substantially independent of the magnification of the optical relay. Thus, for a given beam diameter and deflection angle at the scanning mirror, an increase in magnification at the pupil results in a smaller deflection angle at the pupil.

[0009] The dilated diameter of the human pupil can reach 8 mm, and the effective focal length is approximately 17 mm. The eyes of mice are smaller, with a pupil of 2 mm and an effective focal length typically of about 1.9 mm. For the same deflection angle, a smaller scanning mirror can be used for small animal subjects.

[0010] Webb (1987) described a scanning system that had a multifaceted mirror rotating at 37,800 rpm, which included 25 facets of 6 mm that were overfilled by the returning beam from a dilated 8-mm diameter human eye pupil. He also described another unit magnification configuration with a 5-mm diameter galvanometer mirror driven in a resonant manner, which likewise failed to capture all of the light returning from the pupil. For slower vertical scans, a larger galvanometer mirror located at a separate pupil conjugate was employed.

[0011] A high frame rate is desirable. Animal subjects are typically anesthetized during ophthalmic examinations, but this does not suppress all eye movements, especially those associated with the subject's heartbeat. To reduce motion blur, individual images should be captured over a short enough time period to minimize the resolution loss due to such movements. Frame rates below 15 Hz are problematic, and 30 Hz or higher is preferred.

[0012] To sample 500 lines across the field of view using a simple raster scan at 30 Hz, a line rate of 15000 Hz is required. A rotating polygon mirror can achieve this while maintaining a constant deflection rate of the beam in one direction. A second deflecting mirror is required to scan in the orthogonal direction, ideally located at the optical conjugate of the polygon facets. A mirror driven by a galvanometer near its resonance frequency is an alternative, but in existing instruments, a second mirror is required to perform two-dimensional scanning at a line rate higher than 10000 Hz. Instruments constructed along these lines are bulky and not convenient for use with small animals.

[0013] A simpler and more compact optical system can be achieved if a single mirror deflects the beam in two orthogonal directions. The present invention recognizes that for the small diameter of the mouse pupil, using such a single reflecting surface for two-dimensional scanning is a practical approach. In particular, MEMS (Micro-Electro-Mechanical Systems) technology is used to fabricate such a mirror for an optical projector, typically employing photolithographic patterning and etching of silicon, as described in US2010 / 0020379. Such a device can achieve a combination of field of view, beam width at the subject's pupil, scan rate, and optical resolution suitable for small animal SLO. Another advantage of such a MEMS mirror is that its power consumption is much lower than the several watts common for galvanometer scanners.

[0014] A MEMS resonant scanner designed to operate at 7.5 kHz or higher typically has a mirror diameter of about 1 mm or less and is capable of deflecting the scanned beam by 20° to 45°. Using suitable relay optics, a wider deflection angle can be achieved at the pupil, but this comes at the cost of a reduced beam width, which in turn increases the size of the retinal spot due to diffraction. Since a frame rate of 30 Hz or higher limits the number of lines that can be scanned within a single frame, the resolution loss is not a problem for wide-field imaging. In this case, the scan spot, whose width increases proportionally with the field of view, helps to avoid undersampling of the subject and reduce artifacts due to aliasing.

[0015] There is a conflict between the requirement for a wide field of view for navigation and identification of retinal features of interest and the high-resolution imaging for more detailed study of the region of interest. In existing instruments, this problem is solved by providing multiple sets of relay optics to image the scanning mirror into the subject's pupil, each set of relay optics having a different magnification. Changing the magnification requires removing one optical relay unit and replacing it with another. A more convenient arrangement is to change the magnification of the relay optics through an internal mechanism while maintaining a fixed distance between the final element of the tube lens and the subject, restricting the tilt axis of the beam near the subject's pupil, and keeping the scan spot focused on the retina.

[0016] Gray (2017) proposed in US 9 743 831 to use a MEMS scanning mirror with reflective optics and without confocal spatial filtering. Gray pointed out that such a two-dimensional scanning device should provide a scanning deflection of 180 degrees or more, which is impossible for off-the-shelf silicon MEMS components that meet other basic SLO requirements, including mirror diameter, deflection rate, or resonance frequency. Gray proposed a more feasible solution, including a one-dimensional scanning mirror mounted on a coaxial rotating shaft, which can achieve a larger deflection angle in at least one direction, but this will be larger in volume than a solution that relies on MEMS technology to achieve two deflection axes.

[0017] Another aspect overlooked in Gray's US 9 743 831 disclosure is the vergence, or degree of convergence or divergence, of the scanning beam at the subject's pupil. The ability to control the vergence at the pupil is crucial for focusing the scanning beam onto the part of the retina to be imaged. Blaha and Gaida (US 5 071 246, 1991) addressed this problem for human subjects, but it is especially important for small animal subjects because the relatively large ratio of pupil diameter to focal length in small animals limits the tolerable axial defocus. The confocal instrument disclosed in this application uses a limited depth of focus to distinguish image structures at discrete depths within the retina while suppressing signals from adjacent anterior or posterior layers.

[0018] There are significant variations in the refractive error of the eyes of small rodents, not only between individual subjects, but also with the age of individual specimens, as reported by Zhou (2008), and with wavelength and position across the retina, as reported by Geng (2011). To account for this natural variation of approximately 70 diopters, an additional adjustment of 50 diopters is required to adjust the focus from the preganglionic nerve fiber layer to the posterior retinal pigment epithelium, as reported by Lee et al. (2013).

[0019] For small animal imaging, incorporating the scanning optics into a physically compact scanning head greatly facilitates physically rotating the entire scanning head around the subject's pupil to position the region of interest of the retina near the center of the field of view, rather than presenting the animal to a bulky static instrument designed for human subjects.

[0020] Many applications require support for multiple imaging modes. The modes in use include reflectance imaging at visible and near-infrared wavelengths, fundus fluorescence imaging at one or more wavelengths, and autofluorescence at visible or infrared wavelengths. By placing components including an illumination source, a wavelength multiplexer, a filter for separating the excitation signal and the fluorescence signal (wavelength demultiplexing), and detection and control electronics in a separate unit connected to the scan head by a flexible cable and an optical fiber umbilical cable, operational flexibility is provided without compromising the compact form factor and ergonomics of the scan head. The optical umbilical cable is particularly advantageous for technologies such as optical coherence tomography (OCT) or fluorescence lifetime spectroscopy, where the detection equipment or illumination source may be much larger than a simple semiconductor detector or laser diode.

[0021] While confocal imaging can distinguish different layers in the retina, the axial resolution is limited by the numerical aperture (NA) at the retina, which is approximately equal to the beam radius at the pupil divided by the effective focal length of the subject's eye. For a smaller beam radius, the axial resolution is inversely proportional to the square of the NA. For a wider beam, the theoretical resolution is further degraded by the optical aberrations of the subject's eye. Optical coherence tomography (OCT) provides a useful improvement in axial resolution, which is independent of the beam width, as explained by Tomlins and Wang (2005). The application of OCT in ophthalmic imaging is increasing.

[0022] OCT uses a broad-spectrum light source with low temporal coherence and mixes the light reflected from the subject with light from a reference optical path. Strong coherent interference between the two beams of light can only be observed when the difference in the optical path lengths of the two beams is less than the coherence length of the light source. Ophthalmic OCT typically uses a depth-first scanning mode, where axial scans (A-scans) are sequentially recorded for each point of interest along a scan line to create a two-dimensional B-scan. Volume scans are captured by recording a series of parallel B-scan slices. The A-scan rate is relatively slow, with a repetition rate typically less than 100 kHz, and is not compatible with the high-speed resonant scanning mirrors required by the present invention.

[0023] Podoleanu and Jackson describe in US 5 975 697 an en-face imaging system where phase or frequency modulation is applied to the reference optical path such that coherent interference and detection result in an oscillating electrical signal. Bandpass filtering and demodulation produce a signal that indicates the light power reflected from the subject at the axial depth corresponding to the current reference path length. This requires relatively high-speed detection electronics compared to more traditional SLO operation, but allows two-dimensional en-face scans to be recorded at the same frame rate as SLO scans with the same lateral resolution. Depth information is recovered by changing the length of the reference optical path, typically by moving a reference reflector, and thus capturing consecutive frames at different depths within the subject's eye.

[0024] The advantage of front scanning is that the beam width at the pupil, which optimizes the lateral resolution, is used to adjust the retinal focus to match the depth of front scanning. This reduces the axial depth of focus. In contrast, traditional depth-first A-scan methods tend to have a smaller beam width with poorer lateral resolution, especially when using Fourier-domain or swept-source OCT.

[0025] If the reference signal is split into two parallel paths that differ in length by a quarter of the center wavelength of the light source, high-speed modulation and detection are not required. The light reflected from the subject is also split in two differential interferometers and mixed with the reference beam such that their outputs represent the quadrature components of the beat signal. This technique is widely used in coherent optical communication systems at high gigahertz frequencies but has also been used by Adie et al. (2009) for optical coherence elastography at audio frequencies.

[0026] Tissue is generally birefringent, causing the state of polarization of light to evolve during propagation. In traditional OCT, only the return signal components in the polarization state that is the same as the polarization state of the reference signal contribute to the detected signal, reducing the image intensity from some regions of the subject. By detecting the signals returning in each of two orthogonal polarization states and summing the contributions, polarization fading can be avoided. Commercial modules such as the Thorlabs INT-POL-1300 provide this capability.

[0027] Polarization-sensitive optical coherence tomography uses tissue birefringence to image tissue structure, as described by de Boer (2017). By applying the above orthogonal detection scheme to the two orthogonal polarization components of the return signal, the intensity and polarization state of the return signal can be determined from four simultaneous measurements of the in-phase and quadrature fields in the two polarization states. This both eliminates polarization fading and enables front polarization-sensitive OCT of the subject. SUMMARY OF THE INVENTION

[0028] The present invention discloses a scanning laser ophthalmoscope for examining the fundus or retina of small animals. The scanning head includes a scanning mirror fabricated as a microelectromechanical systems (MEMS) component, having means for tilting about two orthogonal axes and typically etched from silicon. In a preferred embodiment, one axis of the scanning mirror is driven to excite mechanical resonance to achieve a wide deflection angle with low power consumption.

[0029] The use of a single scanning mirror simplifies the optical design. A device for rapidly adjusting the axial focus of the instrument is provided, as well as an optical relay for effectively coupling the scanning beam into the eye of the subject and for the collection and confocal spatial filtering of the light returning from the fundus. Preferably, the optical relay allows the user to scale or change the field of view at the subject, enabling imaging of the area of interest at a higher magnification and improving the optical resolution and signal-to-noise ratio. By combining the fast focusing capability with confocal spatial filtering, axial slicing can be performed to extract three-dimensional information. These devices are combined into a compact unit, which has significantly improved performance and ergonomics compared to existing instruments for small animal research.

[0030] In a preferred embodiment, a flexible optical interconnect device, such as an optical fiber cable, transmits the reflected or fluorescent emission from the subject to the detection system. The operational flexibility is enhanced if the excitation illumination is transmitted through a flexible optical waveguide that carries a single spatial mode from one or more selectable light sources. An optical coherence tomography module employing orthogonal detection can enhance the axial resolution and slicing ability, thus enabling en face imaging at a frame rate and field of view similar to other imaging modalities.

[0031] According to one aspect of the present disclosure, there is provided an ophthalmic imaging device including an optical coherence tomography device. The optical coherence tomography device includes: a light source having low temporal coherence; a beam splitter for simultaneously directing the output from the light source to a subject and a reference optical path, the reference optical path including a delay module for changing the optical path length of the reference path optical path; and an optical combiner for combining the light reflected or scattered from the subject with the light transmitted through the reference optical path, wherein in-phase and quadrature interference beat signals are generated by a quadrature detector including 90-degree optical mixing and balanced optical detection.

[0032] According to another aspect of the present disclosure, there is provided an ophthalmic imaging device including an optical coherence tomography device, the optical coherence tomography device further including: a light source having low temporal coherence; means for simultaneously directing the output from the light source to a subject and a reference optical path, the reference optical path including means for changing the optical path length of the reference path optical path; means for combining the light reflected or scattered from the subject with the light transmitted through the reference optical path, wherein in-phase and quadrature interference beat signals are generated by a quadrature detector including 90-degree optical mixing and balanced optical detection.

[0033] According to yet another aspect of the present disclosure, there is provided a method for imaging the fundus of a small animal (such as a zebrafish or a rodent, such as a mouse), including using the ophthalmic imaging device according to the first aspect or other aspects of the present disclosure.

[0034] In a further aspect of the present disclosure, there is provided an instrument including an optical scanning head, the scanning head including: an optical excitation source having high spatial coherence; a first optical system for guiding light from the optical excitation source onto a scanning mirror; a variable focusing optical device for changing the divergence of the excitation light guided by the first optical system onto the scanning mirror; a second optical system for forming an image of the scanning mirror near the pupil of the subject's eye such that the excitation light is focused within or near the retina of the subject's eye; a beam splitter for separating the returned light from the path of the excitation beam after the light returned from the retinal focus is reflected by the scanning mirror; and a lens for focusing the returned light onto a point conjugate to the retina; a spatial filter located at the retinal conjugate; and a flexible coupler for transmitting the light transmitted by the spatial filter to a detection system. The scanning mirror may be part of a microelectromechanical system including: the scanning mirror capable of tilting about a first axis substantially aligned with a first diameter of the scanning mirror; a driver for exciting the scanning mirror to resonate and oscillate about the first axis; and an actuator for tilting the scanning mirror about a second axis, the second axis being substantially aligned with a second diameter of the scanning mirror and orthogonal to the first axis.

[0035] In a further aspect of the present disclosure, there is provided an instrument including an optical scanning head, the scanning head including: an optical excitation source having high spatial coherence; a first optical system for guiding light from the optical excitation source onto a scanning device including a scanning mirror; a focusing device for changing the divergence of the excitation light guided by the first optical system onto the scanning mirror; a second optical system for forming an image of the scanning mirror near the pupil of the subject's eye such that the excitation light is focused within or near the retina of the subject's eye; a device for separating the returned light from the path of the excitation beam and focusing it onto a point conjugate to the retina after the light returned from the retinal focus is reflected by the scanning mirror; a spatial filter device located at the retinal conjugate; a flexible coupling device for transmitting the light transmitted by the spatial filter to a detection device; wherein the scanning mirror is part of a microelectromechanical system including: the scanning mirror capable of tilting about a first axis substantially aligned with a first diameter of the scanning mirror; a device for exciting the scanning mirror to resonate and oscillate about the first axis; a device for tilting the scanning mirror about a second axis, the second axis being substantially aligned with a second diameter of the scanning mirror and orthogonal to the first axis.

[0036] It will be understood that features described in relation to one aspect of the invention may be incorporated into other aspects of the invention. For example, the apparatus of the first aspect of the invention may incorporate any features described for the apparatus of other aspects of the invention, and vice versa. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of an optical configuration, signal processing, and control system.

[0038] Figure 2 It shows a collimated beam projected through the pupil of a subject's eye, focused on the retina, and the scattered light leaving the pupil substantially parallel to the incident beam.

[0039] Figure 3 It shows an inclined beam entering the pupil, where the light scattered from the retinal patch leaves the pupil substantially parallel to the incident illumination.

[0040] Figure 4 It shows how the vergence of the light incident on the pupil determines the depth of focus of the light within the retina.

[0041] Figure 5 It shows a 4-f optical relay for imaging a scanning mirror into the pupil of a subject.

[0042] Figure 6 It shows a variable magnification optical relay in a low magnification configuration, where the beam has a wide angular deflection at the subject.

[0043] Figure 7 It shows a variable magnification optical relay in a high magnification configuration, where the beam has a reduced angular deflection at the subject.

[0044] Figure 8 It shows an embodiment of a detection system with two optical detectors that respond to different wavelengths selected by a reconfigurable filter.

[0045] Figure 9 It is a schematic diagram of an optical zero-crossing detector for synchronizing the time of signal sampling and digitization with the deflection of a scanning mirror.

[0046] Figure 10 It shows the internal configuration of a compact scan head implemented using an optical path folded around a rigid chassis.

[0047] Figure 11 It shows a control unit and a table that provide a stable and adjustable mounting for the scan head and a holder to support a small animal subject.

[0048] Figure 12 It shows an excitation source capable of providing illumination at multiple wavelengths for fiber-coupled optical coherence tomography.

[0049] Figure 13An embodiment of optical coherence tomography is shown that uses phase modulation of a reference path and bandpass filtering and demodulation of an electrical beat signal to support forward imaging.

[0050] Figure 14 An alternative OCT reference path and demodulation apparatus using acousto-optic modulation, balanced detection, and phase-sensitive recovery of a coherent beat signal are shown.

[0051] Figure 15 An embodiment of forward OCT is shown that uses a 90-degree optical hybrid and two balanced detectors to recover in-phase and quadrature beat components.

[0052] Figure 16 A preferred embodiment of forward OCT using a 90-degree optical hybrid is shown, where the OCT probe is coupled to the scan head using a dedicated fiber optic coupling port.

[0053] Figure 17 A polarization diversity OCT detection scheme using a 90-degree optical hybrid with a balanced detector to detect two orthogonally polarized states of a signal is shown. DETAILED DESCRIPTION

[0054] The invention is first described in an overview in reference Figure 1 and then described in more detail. Light from excitation source 100 is guided by flexible waveguide 11 to a fiber optic coupling assembly, where lens 12 of the fiber optic coupling assembly guides a substantially collimated light beam through beam splitter 14 to focusing device 21 and first optical relay system 20. Lenses 23, 24, 25 of the first optical relay system 20 project an image of aperture 22 onto scanning mirror 31 driven by MEMS module 30 after reflection by steering mirror 32.

[0055] A second optical relay system 40 projects an image of the scanning mirror into pupil 51 of subject eye 50. Light emitted from and exiting pupil 5 from the scan spot is focused back onto scanning mirror 31 and retraces the path of the excitation beam as far as beam splitter 14, where it is deflected into a fiber optic coupler, and lens 62 of the fiber optic coupler focuses the returned light onto aperture 64. The aperture may include the core of fiber optic waveguide 61 that transports the captured light to detection system 70.

[0056] The scanning mirror and associated optics are contained within a compact scan head that is connected to a base unit including excitation source 100, detection system 70, signal processing module 220, and control module 210 by optical umbilical 201 including fibers 11, 61 and electrical umbilical 202.

[0057] The detector output is digitized by the signal processing module 220 and transmitted to the control module 210 that communicates with the host 300. The host 300 in turn forwards user commands to the instrument and displays images from the instrument on the screen 301. The control module 210 sends configuration commands to the instrument subsystems, which include but are not limited to the excitation source 100, the focusing device 21, the driver of the MEMS module 30, the zoom actuator 49 of the second optical relay 40, and the optical filter selection in the detection system 70.

[0058] The present invention will now be described in more detail. A scanning laser ophthalmoscope (SLO) projects a substantially collimated beam into a subject's eye, where the beam is focused on the retina. Light scattered, reflected, or generated by fluorescence is emitted from a small illuminated volume and exits the subject's pupil. The instrument intercepts and measures a portion of this light. A significant feature of the SLO is that the light mainly only returns from those parts of the eye that are illuminated by the scanned beam, which improves the image contrast compared to a conventional fundus camera that floods the retina with light and collects light from all parts of the eye simultaneously.

[0059] Figure 2 A collimated beam 1 is shown incident through the pupil 51 of a subject's eye 50 and focused onto a small spot 5 on the retina 52. The light 2 emitted from the spot 5 and exiting the pupil is substantially collimated and exits parallel to the incident beam. Figure 3 It is shown that when the incident beam 1 is tilted by the illumination system, the spot 5 is focused at different points on the retina, but the returned light 2 remains parallel to the incident illumination and can thus be intercepted by the scanning mirror.

[0060] The contrast is further significantly enhanced by confocal spatial filtering of the returned beam. Figure 1 The spatial filter shown can be a physical aperture 64, or the core or mode field of an optical waveguide 61, and is arranged to be confocal with the scanned spot on the retina. In the absence of scattering along the optical path, only light emitted from or very close to the scanned spot reaches the detector. In practice, some scattering is inevitable, but as the distance between the scattering point and the focused point on the retina increases, the proportion of scattered light reaching the detector rapidly decreases.

[0061] Selectively collecting light emitted near the focal plane can generate images of structures at different depths within the retina. To take advantage of this, a device for changing the focus is needed to accommodate differences in refractive error between subjects and within an individual subject, to accommodate differences with age, across the retina, and for different wavelengths of light.

[0062] Figure 4Shows the effect of changing the vergence of the scanning beam at the pupil 51. Vergence is a measure of the intensity of the illumination converging towards or diverging from the surface of interest, and is quantified by the reciprocal of the distance from the divergence or convergence point to the surface. For a diverging beam 3, the vergence is negative and the focal point 53 moves towards the back of the eye. For a converging beam 4, the focal point 54 moves towards the front of the eye. Preferably, the incident beam remains centered at the pupil 51 and the width at the pupil remains substantially constant as the vergence changes.

[0063] The retinal thickness of a mouse eye is typically about 0.2 mm, the effective focal length is 1.9 mm, and the refractive index of the vitreous medium is 1.34. For a total vergence range of 120 diopters, a change range of about 50 diopters (50 m -1 ) at the pupil is required to change the focal point through the retinal thickness, and a change range of 70 D to accommodate differences across the retina and between different subjects.

[0064] Focusing can be achieved using conventional glass or polymer optics and means for adjusting the position of one or more elements. Ideally, the response time should be comparable to or shorter than the period of a single video frame (typically 30 ms) to enable the capture of fast image sequences at different axial depths within the retina. In a preferred embodiment, a lens element with a variable focal length is used. This can include an electrowetting lens, such as the Varioptic lens type produced by Corning; a deformable lens, such as the deformable lens produced by Optotune, which uses electromagnetic actuation to change the curvature of a transparent membrane hydraulically, or other components with equivalent functionality.

[0065] The instrument described by Blaha (US 5 071 246) has two independent focus adjustments for the incident light and the reflected light respectively. In Figure 1 the instrument shown, the focusing element is shared by the excitation light path and the detection light path, where the return signal is separated by a beam splitter 14 after passing through the focusing device 21. This ensures that in reflection imaging, an axial focus change of the beam incident on the subject results in the same change in the return signal, such that the back-propagating beam has the same vergence at the beam splitter.

[0066] The spatial resolution of the instrument mainly depends on the spot size focused on the retina. A smaller spot size is most easily achieved if the light has spatial coherence over the emission area. In a preferred embodiment, the excitation source is delivered by an optical waveguide (such as a single-mode fiber) that supports a single transverse optical mode.

[0067] Semiconductor lasers and superluminescent light-emitting diodes (SLEDs) are convenient light sources, where the active region is an optical waveguide, enabling the mode field to have high spatial coherence and be efficiently coupled into polarization-maintaining fibers. More generally, fiber coupling allows the use of larger-volume supercontinuum light sources or mode-locked lasers without sacrificing the ergonomic advantages of a compact scanning head.

[0068] It is advantageous to support multiple excitation wavelengths and provide the ability to simultaneously employ different imaging modes. For example, fluorescence imaging excited by blue wavelengths of 480 nm or shorter is combined with reflectance imaging at near-infrared wavelengths such as 785 nm.

[0069] Multiple laser sources can be combined using lenses for collimating the outputs of the multiple laser sources, dichroic filters for combining the collimated outputs, and final lenses for coupling the multi-wavelength beam into the excitation fiber. A fused-tapered fiber coupler is a viable alternative, which has low insertion loss and does not require careful alignment of the combined beams. Using suitable equipment, splicing pigtail laser sources to the fiber coupler is straightforward and generally has lower insertion loss than fiber connectors relying on precision ferrule alignment.

[0070] When the excitation source covers a wide wavelength range, the issue to consider is to ensure single-mode operation at the shortest wavelength while maintaining acceptable bending tolerances of the flexible interconnection at the longest wavelength. An optical waveguide is characterized by its normalized frequency or V number, which is defined as:

[0071]

[0072] where λ is the operating wavelength, a is the radius of the core, n1 is the peak refractive index of the core, and n0 is the refractive index of the cladding. Theoretically, when V is less than 2.405, a fiber with a cylindrical core and a step-index profile supports only two polarizations of a single transverse mode. In the weak-guidance approximation, the linearly polarized fundamental mode is designated as LP01. When the V number is greater than 2.405, additional LP11 modes are supported. This is undesirable and leads to instability in the mode field and the shape and position of the spot focused on the subject's retina.

[0073] For smaller V numbers, especially those less than 1.5, the fiber becomes increasingly sensitive to bending in the waveguide. If the refractive index difference between the core and the cladding is increased, the bending performance may be significantly improved, but this requires a smaller core to maintain single-mode operation, resulting in lower coupling efficiency and reduced tolerance to fiber connector misalignment.

[0074] In a preferred embodiment, the pump fiber is single-mode and has a normalized frequency greater than 1.5 at the longest pump wavelength. At the shortest wavelength, the fiber may no longer be strictly single-mode. However, the LP11 mode is anti-symmetric and its mode field amplitude goes to zero at the center of the core. When coupling into or between fibers, precise coaxial alignment between the cores can minimize the excitation of these LP11 modes. Excitation of higher-order LP02 and LP21 modes can be avoided by ensuring that the normalized frequency V is less than the theoretical cut-off value of 3.83 or possibly higher for an actual refractive index profile.

[0075] Higher-order modes can be further suppressed, for example, by locally stretching and thinning the fiber core such that the V number is reduced below 2.4 and LP11 and higher-order modes radiate into the transparent cladding medium. Such a mode filter can be fabricated using equipment for fusion splicing or during the manufacture of a fused fiber coupler.

[0076] In Figure 1 the preferred embodiment shown, the output from the pump fiber 11 is intercepted by a lens 12 and directed as a substantially collimated beam into a focusing device 21. The intensity distribution of the beam approximates a Gaussian distribution. To prevent unwanted scattering, especially when the beam width at the scanning mirror extends beyond the central mirror reflection area of the scanning mirror, the beam is truncated by an aperture 22 that forms part of or is positioned adjacent to the focusing device.

[0077] Effective coupling from the aperture to the scanning mirror is achieved by a first optical relay 20 that projects a virtual image of the aperture onto the scanning mirror 31. Typically, two or more lenses or lens groups 23, 24, 25 are required to control the magnification and divergence of the image of the aperture on the scanning mirror. By positioning the aperture and the scanning mirror in conjugate positions, the size of the beam incident on the scanning mirror is independent of the beam divergence variation caused by the focusing device. To minimize stray light scattering from the scanning mirror, the magnification of the optical relay is chosen such that the diameter of the virtual image of the aperture is slightly smaller than the central reflective portion of the scanning mirror.

[0078] Those skilled in the art will appreciate that plane (i.e., flat) steering mirrors can be introduced at one or more points along the path from the lens 12 to the scanning mirror 30, thereby folding the optical path and reducing the overall size of the scanning head.

[0079] In the case where there are no optical aberrations in the optical instrument and the subject's eye, the optical resolution and signal-to-noise ratio are optimal when the width of the scanning beam at the subject's pupil is as large as possible, which is consistent with avoiding vignetting, whereby the outer part of the beam is intercepted by the iris or otherwise cannot reach the retina. In this case, the diameter of the diffraction-limited spot that can be focused onto the retina is inversely proportional to the diameter of the beam at the principal surface of the lens and contributes to the theoretical limit of the resolution available for imaging fine structures in the retina.

[0080] In practice, the optical aberrations of the subject will broaden the retinal spot to an extent that increases rapidly with the increase in the beam diameter. The typical pupil diameter of a mouse is 2 mm. Zhang (2015) found that the lateral resolution is optimal when the beam width is 1.3 mm, which is roughly consistent with the simulations based on Zernike aberration coefficients reported by Geng (2011) in an earlier publication. For a narrower beam, optical diffraction broadens the spot. With a wider beam, the resolution is reduced by the increase in optical aberrations.

[0081] Compared with non-resonant devices, resonant scanning mirrors increase the rate of the scanning lines but limit the size of the mirror for acceptable vertical resolution and frame rate. For a vertical resolution of 320 lines at 30 frames per second, the ideal resonant frequency is at least 5 kHz. For higher vertical resolutions of 620 lines at 30 Hz or 720 lines at 25 Hz, the resonant frequency is preferably 10 kHz or higher. MEMS mirrors with this performance have been developed for projection applications. Examples include modules with a diagonal field of view of 50° and a mirror diameter of 1 mm.

[0082] Scanning mirrors with resonant deflection in two axes are available and trace a Lissajous scanning pattern rather than a regular raster. For a given frame rate and number of lines, compared with sinusoidal deflection, a uniform linear or stepped ramp in the vertical direction results in a slower peak scanning speed, a longer integration time, and a better signal-to-noise ratio at the center of the field of view and is thus preferred. US2010 / 0020379 describes a hybrid scanning mirror having a horizontally resonant electrostatically driven axis and an electromagnetically driven vertical axis in a gimbal configuration such that the axes of rotation are coplanar. The resonant oscillation of the inner gimbal provides a high line rate, while the electromagnetic drive of the outer gimbal can perform precise stepped deflections to maintain a uniform line spacing over the entire vertical deflection range.

[0083] We require that the light incident on the scanning mirror be projected into the eye and onto the retina of the subject, and that most of the light returning from the subject be captured by the detection system. In a preferred embodiment, this is achieved by a second optical relay system 40 that projects a virtual image of the scanning mirror into the pupil of the subject's eye such that the scanning mirror and the pupil of the subject are in conjugate positions.

[0084] For a well-corrected optical system, the product of the solid angle swept by the scanning mirror and the cross-sectional area of the beam measured perpendicular to the propagation direction is constant. Ignoring eye aberrations, the optical resolution is maximized when the beam fills the pupil. However, magnifying the image of the scanning mirror at the pupil reduces the sine of the angular deflection of the scanning beam by a factor equal to the pupil magnification. A lower transverse magnification can be used to increase the field of view. This results in a smaller beam width at the pupil of the subject. In this case, the lower resolution and larger diffraction-limited spot size at the retina reduce aliasing artifacts that would otherwise occur if the angular separation of consecutive lines exceeds the optical resolution.

[0085] The results of Zhang (2015) indicate that for typical mouse subjects, eye aberrations are the limiting factor. The transverse optical resolution is optimized for a beam width on the order of 1.3 mm at the pupil of the subject. Independent simulations using aberration coefficients published by Geng (2011) show that the confocal spatial-filtered signal power peaks for beam widths between 0.6 and 1.6 mm. The exact optimal value depends on the size of the confocal spatial filter, the intensity distribution of the incident illumination beam, and the magnitude and characteristics of the optical aberrations in the subject. In general, when the proportion of the returning light is low, a relatively large confocal spatial filter combined with a larger beam diameter maximizes the capture efficiency. Higher transverse resolution can be obtained through strict spatial filtering such that the conjugate image of the spatial filter is comparable to the diameter of the diffraction-limited Airy disk. A larger beam diameter can improve the axial resolution, but there is little benefit for beam diameters greater than 1.6 mm.

[0086] For the expected range of eye aberrations, a beam width of 1.0 to 1.4 mm enables good resolution and good axial resolution when the focus is varied. Using current MEMS technology, the diameter of a biaxial scanning mirror with a resonance frequency of up to 10 kHz is relatively small, approximately 1 mm, with a diagonal field of view typically of 50° or less. Rotation of the scanning mirror off-normal incidence reduces its effective diameter, which is measured in a plane orthogonal to the beam direction. Additionally, to avoid the insertion loss introduced by a beam splitter, the preferred embodiment employs an off-axis rotating mirror 32 that increases the worst-case angle of incidence and further reduces the maximum beam width at the scanning mirror. For an angle of incidence less than 25° at the scanning mirror, the beam width can be 90% of the mirror diameter provided its center is precisely on the axis of rotation.

[0087] If the user can control the magnification of the optical relay 40, a seamless transition can be achieved from a wide field of view for navigation and initial assessment to a narrower region of interest with high lateral and axial resolution.

[0088] Figure 5 A 4-f afocal relay system is shown that has two lenses 41 and 42, the distance between whose principal planes is equal to the sum of their focal lengths. The scanning mirror 31 is located in the front focal plane of the lens 41, and the optical relay forms an image 39 of the scanning mirror at the rear focal plane of the lens 42. Similar fixed magnification configurations are widely used in other instruments and ensure that the image of the scanning mirror remains centered on the subject's pupil, unaffected by the deflection of the scanning beam and the adjustment of the beam divergence for focusing.

[0089] Figure 6 A modified configuration is shown that has an additional lens 44 inserted between the scanning lens 43 and the tube lens 45. The axial position of the lens 44 can be changed. When close to the tube lens 45, the pupil magnification is less than 1, but the sine of the deflection angle increases inversely with the magnification.

[0090] Figure 7 The behavior is shown with the lens 44 now positioned closer to the scanning lens 43, which results in a greater magnification of the image 39 in the subject's pupil, but a correspondingly smaller deflection angle and field of view. Generally, the lens 44 has two positions that provide the same image position but different magnifications. At an intermediate position, the image moves closer to the relay. Depending on the magnification range, this image displacement may be acceptable. To suppress the image displacement, the lens 44 is replaced by two or more lens elements whose spacing can be changed such that the combined refractive power also changes. With two degrees of freedom, the scanning mirror conjugate to the subject's pupil can be stably positioned within a certain magnification range.

[0091] In a preferred embodiment, changing the magnification of the optical relay between the scanning mirror and the subject does not change the distance between the scanning head and the subject and maintains the position of the scanning beam tilt axis within the subject's pupil.

[0092] It will be apparent to those of ordinary skill in the art that when using a single-element lens, the relay system will introduce chromatic aberration and other aberrations. If each of the lenses 43, 44, 45 is replaced with a lens group comprising elements having complementary dispersion (refractive index varying with wavelength) and surface curvature such that the position of the beam tilt axis in the pupil is stabilized, the spot focused on the retina is substantially aberration-free and lies in the same focal plane for the various magnifications supported and within the excitation and detection wavelength ranges of interest, these geometric aberrations can be greatly suppressed. Computer simulation and optimization of detailed lens parameters to meet specific performance requirements is common practice and is supported by software such as Zemax OpticStudio or Synopsys’ Code V.

[0093] In an alternative embodiment, the lens 44 is a negative refractive power diverging lens group rather than a converging lens. In the case where the gaps between the scanning mirror and the scanning lens 43 and between the tube lens 45 and the subject are comparable, this combination can provide the same wide field of view at the subject with a smaller maximum diameter of the zoom element 44. The drawback is that the outer elements 43 and 45 require higher light-gathering refractive power, such that for the same degree of aberration, their design may be more complex and the overall length of the optical relay assembly is longer than that of a positive zoom group.

[0094] By way of example only, potential combinations of beam width and field of view suitable for use with a scanning mirror having a 1 mm diameter and a 50° diagonal field of view are:

[0095]

[0096] For the maximum scan angle at the lowest magnification, there is a risk of vignetting unless the zoom group and the tube lens diameter are large. Depending on the performance requirements, it may be advantageous to use smaller diameter lens elements, resulting in a smaller but more circular field of view and sacrificing the extreme corners of the rectangular scan of the scanning mirror. In addition to cost and weight considerations, smaller diameter lenses can make the optical relay more compact and enhance visibility and accessibility to the subject.

[0097] In a preferred embodiment, the beam splitter 14 separates the incident beam and the return beam, rendering both substantially collimated. In one embodiment, the collimated beam is focused onto a hole 64 conjugate to the scan spot at the retina, and the light transmitted through the spatial filter hole is coupled into a multimode fiber, as described by Blaha. In a preferred embodiment, a precision fiber coupler focuses the light onto the core of a few-mode or multimode fiber such that the fiber core constitutes the spatial filter.

[0098] In the absence of aberrations, the transverse resolution of the confocal SLO is increased by a factor of 1.4, with single-mode input and matched single-mode output, such that the image of the mode field at the retinal conjugate has the same size and position for each fiber. In practice, aberrations broaden the image spot, and a matched mode-field configuration has poor light collection efficiency and reduces the signal-to-noise ratio. For typical aberrations, an effective compromise is to choose the core diameter of the collection fiber such that its image at the retina is a fraction of the diffraction-limited Airy disk diameter from a uniformly illuminated pupil with the same diameter as the incident scanning beam. A multiple between 1 and 5 times the Airy disk diameter provides good collection efficiency while maintaining acceptable transverse and axial spatial resolution.

[0099] To maximize the coupling efficiency, the numerical aperture of the detection fiber should not be less than that of the return beam focused onto the fiber by the lens 62 of the fiber coupler.

[0100] Just by way of example, commercially available single-mode fibers typically have a numerical aperture in the range of 0.1 to 0.14. A fiber with a cut-off wavelength of 600 nm will have a core diameter of approximately 4 μm and exhibit low bending losses for excitation wavelengths up to 780 nm or 830 nm. As mentioned above, shorter wavelengths around 480 nm can be used with precise centered light source coupling or appropriate mode filtering. If both excitation 12 and detection 62 use coupling lenses with similar focal lengths, a spatial filter 64 or the core diameter of the fiber 61 in the range of 8 to 25 μm can be used. For example, telecommunication fibers with a single-mode cut-off wavelength of 1250 nm or 1400 nm have a core diameter of 9 - 10 μm. For imaging modes such as autofluorescence, the return signal is typically low, and a multimode fiber with a core diameter of 50 μm and a numerical aperture of 0.2 will collect more light and provide better signal-to-noise ratio performance with only a small loss in resolution.

[0101] A potential problem with such instruments is the Fresnel reflection at the interface between the beam splitter and the subject (wherever there is a change in the refractive index of the propagation medium). There are well-known mitigation techniques to minimize the power reaching the detector from such reflections. In any case, it is recommended to coat the air-glass interface with a dielectric antireflection coating to improve the transmittance. For planar and large-radius surfaces, a small tilt of the angle of incidence away from the normal can deflect the reflected beam away from the confocal spatial filter without introducing significant optical aberrations. A hole or mask can block the reflections from a particular surface, but this becomes difficult when there are numerous surfaces and the ray divergence varies over a wide range with the adjustment of the retinal focus and pupil magnification.

[0102] Fresnel reflection preserves the linear state of polarization. Therefore, an effective means of suppressing reflection is to insert a linear polarizer in the detection beam, whose direction is orthogonal to the polarization direction of the illumination source. The output from a semiconductor diode laser is linearly polarized at the light source and can be transmitted through a polarization-maintaining single-mode fiber, which simultaneously maintains the linear polarization state and the high spatial coherence of the light source.

[0103] In a preferred embodiment, the illumination is provided from a linearly polarization-maintaining fiber via a connecting fiber coupler. Rotational misalignment between the laser diode and the fiber pigtail, as well as polarization cross-coupling in the fiber, may disrupt the polarization state, thereby limiting the degree of polarization extinction that can be achieved by the detector polarizer 63. Therefore, an additional polarizer 13 aligned with the nominal polarization axis of the input fiber is preferred.

[0104] In another preferred embodiment, additional polarization discrimination is provided by making the beam splitter 14 a polarization beam splitter.

[0105] To avoid causing retinal damage, the optical power that can be coupled into the subject's eye is limited as required. The power at the pupil is typically less than 1 mW, especially at shorter visible and ultraviolet wavelengths, because these wavelengths pose a greater risk of photochemical damage to the retina. In imaging modes such as autofluorescence, the returned signal may be very low, and multiple image frames need to be stacked to collect a usable signal. A detector with high quantum efficiency, low noise, and low dark signal is required. Photomultiplier tubes and silicon avalanche photodiodes (Si APDs) have been used. The preferred embodiment uses a multi-pixel photon counter (MPPC) detector, also known as a silicon photomultiplier (SiPM). Compared with typical vacuum tube photomultiplier tubes and Si APDs, the silicon photomultiplier (SiPM) has high avalanche gain, low dark current, and a lower operating voltage.

[0106] Figure 8 A preferred embodiment of the detection module 70 is shown. The light guided by the flexible detector waveguide 61 is collimated by the lens 71 and directed onto the photodetector 81. The bias voltage source 83 controls the avalanche gain, and the transimpedance amplifier 84 converts the current output of the detector into an electrical signal with an output impedance matched to the input of the signal processing module 220.

[0107] To support simultaneous imaging at multiple wavelengths, the dichroic beam splitter 72 redirects one or more bands into a parallel detection path using the second detector 82, which can be selected to have a different spectral response from the first detector 81. Some imaging modes, especially autofluorescence, return very low light levels at the wavelengths of interest, so it is particularly important to reject the excitation wavelengths. Bandpass and edge-blocking filters 74, 75 are inserted according to the requirements of each imaging mode. Preferably, an electromagnetic or other insertion means 77 allows for rapid selection of filter combinations under direct user control or program control. Obviously, an additional parallel detection channel can be added by replacing the mirror 73 with another dichroic beam splitter, such that the transmitted light 76 is processed by a combination of one or more additional filters and detectors.

[0108] In a preferred embodiment, at least one axis of the scan is provided by resonant oscillation of a scanning mirror. The mirror deflection varies sinusoidally with time, and signals are captured during both the forward and reverse scans of the scanning mirror. In principle, the signals can be sampled at a variable rate such that samples are acquired at times corresponding to uniform angular spacing of points on the retina. In principle, an analog-to-digital converter (ADC) with sufficient resolution and bandwidth typically employs a pipelined architecture such that the sampling is synchronized to a constant-frequency clock. In this case, the ADC sample stream must be interpolated to convert the raw data into a rectangular image raster with more uniform pixel spacing.

[0109] It is particularly important to synchronize or otherwise align the signal samples corresponding to the forward and backward mirror scans. If the electrical samples are delayed relative to the scanning mirror oscillation, the image pixels will be shifted behind the angular motion of the scanning mirror and thus move in opposite directions along alternate lines. Pham (2011) describes a capacitive sensor integrated into a microelectromechanical system module for detecting the zero-crossing points at the midpoints of each line. A potential problem with such an on-chip capacitive sensor is noise, resulting in an 86 ns root mean square jitter in the device reported by Pham. The underlying mechanical resonance has higher short-term stability, so the timing error can be reduced by averaging over multiple oscillation periods.

[0110] An alternative is to use an optical zero-crossing detector (OZCD) to track systematic changes in the phase shift and mirror resonance frequency. In a preferred embodiment, as Figure 9Schematically shown, the converging beam 90 formed by the light source 91 and the lens 92 is directed onto the scanning mirror 31, and the reflected beam is focused on the optical detector including the linear sensor element 86. The aperture 93 limits the beam width near the light source and minimizes scattering by ensuring that the beam is confined to the scanning mirror and does not irradiate other parts of the MEMS module 30. The sensor is preferably shielded by a slit 95 which is oriented such that at the zero crossing of the fast resonant scan, when the vertical deflection changes, the slit lies in the plane swept by the OZCD beam, and when the vertical deflection approaches its midpoint, the slit is substantially orthogonal to the reflected beam 94.

[0111] Whenever the OZCD beam 94 reflected from the scanning mirror irradiates the slit 95, the preamplifier 87 and the comparator 88 convert the analog output from the sensor 86 into binary logic level pulses. The pulse width depends on the beam width, the detector or slit width, and the distance to the scanning mirror, and is preferably in the range of 0.05 to 1.0 μs.

[0112] The OZCD pulses are processed by the control module 210. The zero crossing time corresponding to each pulse is preferably calculated as the average of the arrival times of the rising and falling edges. Using the average value greatly reduces the sensitivity to signal level variations caused by laser output power drift or vertical deflection of the scanning mirror.

[0113] The OZCD zero crossing time calculated in this way is still sensitive to the exact lateral position of the slit 95 and its angular alignment. The system delay can be calibrated and corrected in the software or firmware of the control system, but such correction depends on the speed of the scanning mirror and thus also on the deflection amplitude. Ideally, the deflection amplitude can be varied in concert with the variable magnification provided by the second optical relay system 40 under user control in order to optimize the signal-to-noise ratio and image resolution according to the imaging mode used.

[0114] In a preferred embodiment, the time t of the center point of each line is determined by using the weighted average of three or more zero crossings of each scan line C , avoiding the need to recalculate the time correction for each scanning mirror deflection amplitude.

[0115] Generally speaking, the robust estimate of the time t DC at which the scanning mirror oscillation changes direction is given by the average of two consecutive OZCD times, each of which is calculated as the average of the rising and falling edge times of each individual pulse. For a sinusoidal oscillation, this turning point is a quarter cycle ahead of or behind the zero crossing, and thus, the improved estimate of the true zero crossing is:

[0116] t C = t DC ± T / 4

[0117] The oscillation period T is equal to the time between OZCD pulses from consecutive lines scanned in the same direction.

[0118] More specifically, if samples from multiple lines are buffered and the current zero-crossing time t ZC (n), the previous zero-crossing time t ZC (n - 1), and the next zero-crossing time t ZC (n + 1) are saved, then an improved estimate of the nth zero-crossing time is:

[0119] t C (n) = 0.25t ZC (n - 1) + 0.5t ZC (n) + 0.25t ZC (n + 1)

[0120] Alternatively, if the ADC clock phase is adjusted before the start of each line, a robust estimate of the zero-crossing time for line n can be calculated based on the previous three zero-crossing times as follows:

[0121] t C (n) = 1.25t ZC (n - 1) + 0.5t ZC (n - 2) - 0.75t ZC (n - 3)

[0122] A signal processing module 220 is provided, which at least includes an analog-to-digital converter that converts the continuous electrical signal from the detector into a sequence of digital representations of the signal at each sampling moment. Further conversion is required to generate an image suitable for subsequent display or processing by general software tools.

[0123] Samples are acquired during both the forward and reverse scans of the resonant mirror oscillation. The order of the ADC samples from alternate lines must be reversed.

[0124] The delay between the OZCD zero-crossing time and the nearest ADC sampling time corresponds to the spatial displacement between the lines from the forward and reverse scans of the scanning mirror respectively. This can be corrected by interpolation between ADC samples. If the phase of the ADC clock is adjusted with sufficient precision before the start of each line, time alignment by interpolation is not required.

[0125] An ADC with a sufficiently high sampling rate and resolution typically employs a pipelined architecture, which requires a sampling clock of constant frequency. The deflection of the resonant axis of the scanning mirror is a sinusoidal function of time, such that the physical interval corresponding to the center of each line for adjacent ADC samples is larger than that corresponding to the edges. The uncorrected image will be distorted and features will be stretched towards the extreme directions of the (horizontal) resonant axis scan.

[0126] In addition to sine distortion, there is also distortion related to the geometry of the scanning mirror tilt axis and the orientation of the incident light beam relative to the center of the field of view. The general effect is barrel distortion combined with the curvature of the image field away from the direction of the incident light beam. For a beam deflection of ±10° in both the horizontal and vertical directions at the scanning mirror (20°×20° field of view, magnification of 1×), the image distortion is typically less than 5%, but increases with the square of the deflection amplitude.

[0127] Those skilled in the art will understand that there are various means to implement the above distortion correction process, including but not limited to one or some combinations of application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), graphics processing units (GPUs), or microprocessors. One method is to implement minimal processing in the instrument itself and use the host computer 300.

[0128] In a preferred embodiment, time alignment, sine transformation, and geometric projection transformation are applied in the signal processing module 220 and the system control module 210, and the corrected image frames are transmitted to the host computer 300 as discrete images or video streams. The use of industry-standard hardware interfaces (such as Universal Serial Bus (USB)) and mature communication protocols (such as USB Video Class (UVC)) provides the user with the ability to integrate the instrument with existing hardware and software tools.

[0129] Figure 10 An embodiment of the scanning optics is shown, where a relatively long optical path is achieved within a compact scanning head by folding the optical path. The excitation light is delivered from the connecting fiber collimator 15 and is guided through the beam splitter 14. The folding mirror 26 at a 45° angle to the light beam is located between the lenses 24 and 25 of the first optical relay, and the second 45° folding mirror 27 guides the light to the scanning mirror of the MEMS module 30 through the folding mirror 32 (not shown in this view). Two mirrors (not shown) respectively guide the light rays from the OZCD collimator 92 to the scanning mirror and from the scanning mirror to the slit 95 of the OZCD detector.

[0130] Figure 11 The scanning head 10 with its housing installed in place is shown. The translation and rotation adjustment device 9 allows precise control of the position and orientation of the scanning head relative to the bracket 6 provided for supporting and orienting the small animal subject.

[0131] Techniques commonly employed and combined in ophthalmic imaging include: reflection imaging at one or more wavelengths, fluorescence imaging using dyes or autofluorescence (with a certain range of excitation wavelengths), and optical coherence tomography (OCT). If only a limited number of different light sources are required, a single-mode fused fiber wavelength division multiplexing (WDM) coupler is stable and has a low insertion loss. When more flexible wavelength selection is needed, an expanded beam coupler and a dichroic filter are preferred.

[0132] By way of example only, Figure 12 illustrates an apparatus for flexibly supporting a plurality of excitation sources. The outputs from laser sources 101 and 102 are collimated by lenses 111 and 112, the light beams are combined by dichroic filter 120 and coupled into optical fiber 11 through lens 110. For 780 nm and 488 nm light sources, a long-pass filter having a cut-off wavelength between 600 nm and 750 nm is suitable. A second ultraviolet or visible light source 103 can be added through collimating lens 113 and dichroic filter 122. Reflector 123 can be a mirror or a third dichroic filter for supporting a third UV or visible light source and is fiber-coupled through lens 114. In the near-infrared branch, dichroic filter 121 and lens 115 couple the light from optical fiber 430. For optical coherence tomography, the transmission is bidirectional, so that the backscattered light is returned to the OCT module through interconnect 430.

[0133] Figure 13 illustrates an embodiment of OCT. A light source 401 with high spatial coherence and low temporal coherence is required. Superluminescent diode (SLD) light sources can be used in the near-infrared wavelength range, but supercontinuum laser light sources are also used. The achievable axial resolution is inversely proportional to the optical bandwidth of the light source, as described by Tomlins and Wang (2005). For an 850 nm light source with a 50 nm bandwidth (full width at half maximum), the axial resolution in air is approximately 6.4 microns, corresponding to a depth of approximately 4.8 microns in water-containing tissue.

[0134] Directional coupler 402 guides the light from the light source to excitation module 100 through optical cable 430, and excitation module 100 is coupled to scanning head 10 through optical cable 11. Optical relay 40 images the scanning mirror into the eye 50 of the subject. The reflected light and scattered light are transmitted back through the system to directional coupler 402. The second output port of coupler 402 guides the light to a reference path, which includes optical circulator 403, phase modulator 409, collimating lens 405 and mirror 406. Optical fiber cable 404 is selected to match the optical path length through signal cables 430, 11, the scanning optical system and transmitted to the fundus 52 of the subject. Adjustment of the reference optical path length is provided by translation mechanism 408.

[0135] Preferably, optical fiber interconnections 11 and internal OCT interconnections (including 404, 430, 431, 432) are polarization-maintaining and single-mode, supporting orthogonal linear polarization modes within the wavelength range emitted by light source 401.

[0136] The optical circulator 403 guides the backward reflected light from the reference path through the optical fiber 431 to the 50 / 50 directional coupler 420, which preferably distributes the power equally from either of its two input ports to the two output ports. The signal light returning from the subject is coupled through the optical fiber 432 to the second input port. The two output ports respectively output the sum and the difference between the signal light input field and the reference light input field. The photocurrents in the photodiodes 421 and 422 are mainly common mode signals proportional to the sum of the signal light power and the reference light power. The superimposed differential signal is proportional to the square root of the product of the signal power and the reference power. The photodiodes are connected in series in a balanced detector configuration such that the common mode photocurrents cancel each other out. The differential beat current is combined in phase at the input of the transimpedance amplifier 423.

[0137] The amplitude of the beat signal depends not only on the amplitudes of the backscattered light signal and the reference light signal, but also on their respective instantaneous optical phases. The phase modulator 409 is driven to ensure that the optical delay change in the reference optical path is at least one wavelength or more than 360 degrees in phase, so as to generate a differential beat signal with a measurable change in its amplitude. The bandpass filter 424 selects the frequency centered on the modulation frequency, or as discussed by Podoleanu and Jackson in US 5 975 697, the frequency centered on the harmonics of the modulation frequency. The amplitude of the alternating beat frequency is extracted by rectification or other means in the demodulator 425, digitized, and transmitted to the signal processing and control modules 220 and 210.

[0138] The beat signal amplitude is highest when the optical group delays of the signal arm and the reference arm are the same at all optical frequencies. If dispersion causes a wavelength-dependent group delay difference between the two paths, both the beat amplitude and the axial resolution will be reduced. The first-order dispersion correction is carried out by a pair of opposing triangular cross-section glass prisms 407, which allows the effective thickness of the dispersive material in the reference path to be changed without significantly changing the alignment of the expanded beam. For very wide spectral bandwidths, two pairs of compensating prisms can be used to correct higher-order dispersion terms, with each pair using glass with different dispersion capabilities. Kowalevicz et al. (2002) used a first compensator made of BK7 crown glass and a second compensator made of flint glass with higher dispersion.

[0139] The optical circulator 403 ensures that the light is guided through the optical fiber 431 without propagating back to the light source 401 through the directional coupler 402. Such backward reflection can cause optical instability, which in extreme cases can lead to laser oscillation and may cause optical damage. An alternative is to insert an optical isolator between the light source 401 and the coupler 402.

[0140] Figure 14An alternative reference path and detection scheme are shown. The mirror is replaced by a retroreflector or prism 411, and the laterally shifted reflected beam is coupled into a single-mode fiber by a lens 415 and transmitted to an acousto-optic modulator 410. The frequency-shifted reference signal is transmitted through fiber 431 to a 50 / 50 directional coupler 420 and mixed with the backscattered signal from fiber 432. Balanced detection through photodiodes 421 and 422 is used to measure the coherent beat between the reference and the signal, and the differential photocurrent is amplified in a transimpedance amplifier 423. The beat signal is synchronized with the frequency shift caused by the acousto-optic modulator (AOM). Phase-sensitive or coherent detection in an electrical mixer 426 uses the AOM radio frequency (RF) drive signal 425 as a local oscillator to generate a signal proportional to the photodiode beat signal. Preferably, quadrature local oscillator waveforms are generated, and the outputs from the in-phase and quadrature RF mixers are combined so that the output signal is proportional to the root mean square amplitude of the beat signal, independent of the optical and local oscillator phases.

[0141] Figure 15 An embodiment is shown that uses detection techniques common in high-speed optical communication systems, avoiding the need for high-frequency phase or acousto-optic modulation. The signal path and the reference path are similar to those described in Figure 13 but without a phase modulator. The reference fiber 431 and the signal fiber 432 are coupled to the input ports of a 90-degree hybrid 440. For clarity, an optical waveguide embodiment of the hybrid is shown. Each input is equally split by directional couplers 441 and 442, and copies of the signal and reference signals are applied to directional couplers' 443 and 444. The internal interconnected optical path lengths are carefully matched, except for one reference cross-connection where an incremental delay of one-quarter of the period of the center wavelength of the light source is added 448. Balanced detectors 451 and 454 at the outputs of couplers 443 and 444 feed the differential photocurrent into transimpedance amplifiers 450 and 455, and the outputs of the transimpedance amplifiers are proportional to the in-phase and quadrature components of the beat signal. The reflected amplitude is proportional to the square root of the sum of the squares of the two transimpedance outputs:

[0142]

[0143] Although a planar waveguide implementation of the 90-degree hybrid is possible, it would be difficult to maintain an exact quadrature relationship without temperature stabilization. A bulk optical implementation (such as a Michelson interferometer-based design offered by Optoplex Corporation) provides equivalent functionality and is a stable and robust alternative.

[0144] Coupling the OCT detection signal and the return signal through the source multiplexer 100 provides convenience and flexibility. However, the coupling losses of multiple optical fibers to free space, and the insertion losses of the scanning head beam splitter 14 and dichroic filters 110, 115 will reduce the sensitivity and signal-to-noise ratio. Figure 16 A preferred embodiment is shown, in which the OCT illumination is coupled through the collimator 16 and the dichroic filter 17 located between the beam splitter 14 and the focusing lens 21. The filter 17 selectively reflects light within the OCT source wavelength range and transmits other (usually shorter) imaging wavelengths. The same configuration can be used for Figure 13 and 14 the frequency or phase modulator embodiments shown.

[0145] Figure 17 An embodiment of polarization diversity OCT is shown. The light returned from the subject is transmitted through the polarization maintaining fibers 430, 432 to the polarization demultiplexer 460 and split into two orthogonal polarization states, transmitted through the fiber 434 to the 90-degree hybrid 440, and transmitted through the fiber 436 to the second hybrid 449. The reference signal from the delay module 470 is separated by the 3dB coupler 462, wherein the outputs 433 and 435 are coupled to the reference inputs of the two optical hybrids. The four balanced detector outputs from the transimpedance amplifiers 450, 455, 456, 457 are preferably simultaneously sampled by the analog-to-digital converter. Polarization fading is suppressed by calculating the reflection amplitude proportional to the square root of the sum of the squares of the four outputs, and the square root of the sum of the squares of the four outputs is as follows:

[0146]

[0147] The in-phase and quadrature field amplitudes of the two orthogonal polarization states describe the rotation of the polarization state of the signal relative to the reference at the detector, and are the basis of polarization OCT imaging.

[0148] For some forward OCT applications, the translation mechanism 408 for the reflector 411 in the optical delay module 470 may respond too slowly. The variable delay module containing optical fibers provides a faster response, and this variable delay module stretches the optical fiber through the piezoelectric device. By using the matching delay modules 472 and 471 in the signal path 432 and the reference path 431, the axial resolution degradation caused by the fiber dispersion imbalance in the signal path and the reference path can be minimized, and these delay modules are driven partially in antiphase to minimize the difference in the net dispersion at the interferometer.

[0149] References

[0150] US 3 016 000, 1962, Y Sato and K Noyori, "Portable oculi fundus camera"

[0151] US 4 213 678, 1980, O.Pomerantzeff and R.H.Webb, "Scanning laserophthalmoscope for examining the fundus of the eye".

[0152] US 5 071 246, 1991, E.Blaha, G.Gaida, Confocal Scanning Ophthalmoscope.

[0153] US 5 975 697, 1999, A.G.Podoleanu and D.A.Jackson, "Optical mapping apparatus with adjustable depth resolution".

[0154] US 7 993 000 B2, 2011, N.A.Massie, "Method and apparatus for imaging an eye of a small animal".

[0155] US 9 743 831 B2, 2017, D.Gray, D.Swan, M.Thomson, "Retinal imaging apparatus and method".

[0156] US2010 / 0020379 A1, 2010, Yoram Lubianiker, "Gimbaled scanning micro-mirror actuation scheme and architecture".

[0157] U.S. Provisional Patent Application 63 / 380,605, October 24, 2022, Peter West and Alan Robinson, "Scanning Laser Ophthalmoscope for Small Animals"

[0158] S.G.Adie et al., "Audio frequency in-vivo optical coherence elastography", Phys.Med.Biol. Vol. 54 (2009), pp. 3129 - 3139.

[0159] Johannes F. de Boer, Christoph K. Hitzenberger, and Yosihiaki Yasuno, "Polarization sensitive optical coherence tomography - a review", Biomedical Optics Express, Vol. 8, No. 3, 2017, March 2017

[0160] Lee R. Ferguson, James M. Dominguez II, Sankarathi Balaiya, Sandeep Grover, Kakarla V. Chalam, "Retinal Thickness Normative Data in Wild - Type Mice Using Customized Miniature SD - OCT", PLOS, June 27, 2013, DOI: 10.1371 / journal.pone.0067265.

[0161] Ying Geng et al., "Optical properties of the mouse eye", Vol. 2, No. 4, pp. 717 - 738, Biomedical Optics Express, April 2011.

[0162] A.M. Kowalevicz et al., "Ultrahigh resolution optical coherence tomography using a superluminescent light source", Optics Express, Vol. 10, (2002), p. 349.

[0163] D.D Pham et al., "Position Sensing and Electrostatic Actuation Circuits for 2 - D Scanning MEMS Micromirror", Defense Science Research Conference and Expo (DSR), August 3 - 5, 2011, DOI: 10.1109 / DSR17494.2011.

[0164] S. Remtulla and P. E. Hallett, “A schematic eye for the mouse and comparisons with the rat”, Vision Rs. Vol. 1, pp. 21 - 31, 1985.

[0165] Warren J. Smith, “Modern Optical Engineering”, 3rd Edition, McGraw - Hill, 2000.

[0166] A. W. Snyder and J. D. Love, “Optical Waveguide Theory”, ISBN 0 - 412 - 09950 - 0.

[0167] P. H. Tomlins and R. K. Wang, “Theory, developments and applications of optical coherence tomography”, J. Phys. D: Appl. Phys. 38 (2005) 2519 - 2535.

[0168] R. H. Webb and G. W Hughes, “Scanning Laser Ophthalmoscope”, IEEE Transactions on biomedical engineering, BME - 28, No. 7, July 1981, pp. 488 - 492.

[0169] R. H. Webb, G. W. Hughes and F. C. Delori, “Confocal scanning laser ophthalmoscope”, Applied Optics, Vol. 26, No. 8, April 15, 1987, pp. 1492 - 1499.

[0170] D. L. Yan et al., “The phase locked loop for MEMS horizontal scanning control of micro - laser projection ASIC”, International Symposium on Integrated Circuits, 2011.

[0171] Pengfei Zhang et al., "Effect of scanning beam size on the lateral resolution of mouse retinal imaging with SLO", Optics Letters, Vol. 40, No. 24, December 15, 2015.

[0172] Xiangtian Zhou et al., "The Development of the Refractive Status and Ocular Growth in C57BL / 6 Mice", Investigative Ophthalmology & Visual Science, December 2008, Vol. 49, No. 12.

Claims

1. An ophthalmic imaging device, comprising an optical coherence tomography device, wherein the optical coherence tomography device comprises: A light source with low temporal coherence; A beam splitter for simultaneously guiding the output from the light source to a subject and a reference optical path, the reference optical path including a delay module for changing the optical path length of the reference optical path; And An optical combiner for combining the light reflected or scattered from the subject with the light transmitted through the reference optical path, wherein in-phase and quadrature interference beat signals are generated by a quadrature detector including 90-degree optical mixing and balanced optical detection.

2. A method for imaging the fundus of a small animal, comprising using the ophthalmic imaging device according to claim 1.

3. An instrument comprising an optical scanning head, the scanning head comprising: An optical excitation source with high spatial coherence; A first optical system for guiding the light from the optical excitation source to a scanning mirror; A variable focusing optical device for changing the divergence of the excitation light guided by the first optical system to the scanning mirror; A second optical system for forming an image of the scanning mirror near the pupil of the subject's eye such that the excitation light is focused within or near the retina of the subject's eye; A beam splitter for separating the path of the light returning from the retina focus after being reflected by the scanning mirror from the path of the excitation beam after the light is reflected by the scanning mirror, and a lens for focusing the returning light to a point conjugate to the retina; A spatial filter located at the retina conjugate; A flexible coupler for transmitting the light transmitted by the spatial filter to a detection system; wherein the scanning mirror is part of a microelectromechanical system, the microelectromechanical system comprising: The scanning mirror capable of tilting about a first axis substantially aligned with a first diameter of the scanning mirror; and A driver for exciting the resonant oscillation of the scanning mirror about the first axis, and An actuator for tilting the scanning mirror about a second axis, the second axis being substantially aligned with a second diameter of the scanning mirror and orthogonal to the first axis.

4. The apparatus according to claim 3, wherein, The oscillation frequency about the first axis is greater than 5000 Hz, and the diameter of the scanning mirror is less than 2 mm.

5. The instrument according to claim 4, wherein The response time of the variable focusing optical device is less than 30 ms.

6. The instrument according to claim 5, wherein The variable focusing optical device is an electrowetting lens.

7. The instrument according to claim 3, wherein The second optical system includes a zoom actuator for changing the magnification of the image of the scanning mirror formed near the pupil of the subject's eye such that the distance from the image to the second optical system is substantially constant as the magnification changes.

8. The apparatus according to claim 5, wherein, The excitation source is a single-mode or few-mode optical fiber, the distal end of which is coupled to one or more light sources.

9. The apparatus according to claim 8, wherein, The excitation source optical fiber is single-mode at the longest wavelength it supports.

10. The apparatus according to claim 9, wherein, [[ID= 11. The instrument according to claim 3, wherein, ​ A light source that projects a light beam onto the scanning mirror; An optical detector sensitive to light from the light source after reflection from the scanning mirror; and A device for generating a synchronization pulse from the output of the optical detector whenever the angular deflection of the resonant mirror is at an intermediate point between two extremes of its angular deflection range.

12. The instrument according to claim 11, wherein, The synchronization time for each line is calculated based on the times of the rising and falling edges of the optical zero-crossing detector pulses from at least three consecutive scan lines, such that the calculated zero-crossing time is insensitive to small displacements from the exact midpoint of the angular deflection range of the zero-crossing detector.

13. A method for imaging the fundus of a small animal using a confocal scanning laser ophthalmoscope according to claim 5, wherein, The diameter of the light beam projected into the subject's eye is between 30% and 90% of the diameter of the subject's pupil.

14. A method for imaging the fundus of a small animal using a confocal scanning laser ophthalmoscope according to claim 7, wherein, The diameter of the light beam projected into the subject's eye is between 30% and 90% of the diameter of the subject's pupil.

15. An instrument comprising a scanning head according to claim 3, wherein, The excitation source includes an optical coherence tomography device, and the optical coherence tomography device includes: A light source having low temporal coherence; A beam splitter for guiding light from the light source through the scan head to the subject and a reference optical path simultaneously, the reference optical path including a delay module to change the optical path length of the reference optical path; An optical combiner for combining the signal light reflected or scattered from the subject and returned through the scan head with the light transmitted through the reference optical path; A coherent interference detector for detecting the coherent interference between the signal light reflected from the subject and the light from the reference path.

16. The apparatus according to claim 15, wherein, The coherent interference detector includes a plurality of differential optical detectors, wherein the phase delay between the reference path and the signal path differs by a non-integer multiple of the optical period of the central wavelength of the low temporal coherence light source.

17. The instrument according to claim 16, wherein, The coherent interference detector includes a quadrature detector, and the quadrature detector includes a 90-degree optical mixer and two differential optical detectors.

18. The apparatus according to claim 15, wherein, The coherent interference detector includes: A modulator for modulating the optical phase of the light in one of the reference path or the signal path; A detection arrangement for detecting the amplitude of the fluctuations of the coherent interference caused by the modulation of the optical phase of the light in one of the optical paths.

19. The instrument according to claim 18, wherein The modulator is an acousto-optic modulator.

Images