Adaptive Optics (AO) describes the use of wavefront sensors to sense aberrations of ocular optics, and to use deformable mirrors to compensate for the aberrations to enhance retinal imaging performance. While this technique was originally developed to reduce aberrations in Earth's atmosphere when gazing at the night sky, it has been modified and optimized for the visualization of retinal structures in vivo.   Given its utility in retinal imaging, AO is currently under research for optimization in viewing cone photoreceptors (PR), retinal pigmented epithelial cells (RPE), retinal ganglion cells, blood vessels, and the optic nerve, among others. This article will evaluate different types of Adaptive Optics machines, analyzing advantages and disadvantages and the potential applications of each.
AO is often paired with retinal imaging technology to aid in visualization of the posterior segment. AO has been paired with a number of existing ophthalmic imaging modalities including flood illumination ophthalmoscopy (FIO), optical coherence tomography (OCT), and scanning laser ophthalmoscopy (SLO) to reduce the effect of aberrations on retinal imaging. Some of its applications include, but are not limited to, the visualization of cone photoreceptors in healthy eyes, those with age-related macular degeneration (AMD), and those with inherited retinal diseases, the study of vasculature, particularly in patients with diabetic retinopathy, the tracking of leukocyte migration through retinal vasculature, and the imaging of retinal ganglion cells, particularly in patients with glaucoma.  
For AO-FIO the patient sits facing the the machine with his or her chin on the chin rest and forehead braced against a plastic ribbon to steady the head. In order to optimize the consistency of the procedure and to mitigate blinking in the first frame, the patient fixates on a point and initiates the system with a button. The machine begins by rapidly assessing optical aberrations and correcting them with AO until they fall below a designated threshold. At this point, the camera collects an image. This sequence of aberration correction and image collection continues until a sufficient number of images are gathered.   The main drawback of AOFIO remains its image quality, which often exhibits poor contrast due to the capture of scattered light from retina and choroid. 
In AO-SLO, AO runs inseparably from the imaging system, correcting aberrations in real time. One particular feature of this system's deformable mirror is its ability to perform optical sectioning by adjusting the focal plane within the retina, using the defocus control. The defocus control may be built into the closed-loop system, allowing for continuous imaging.  This device functions by focusing a single-spot beam on the retina and reflecting light through a confocal aperture, thought to minimize scatter.  While confocal detection provides excellent imaging of photoreceptors, other detection modes aid in the visualization of other retinal structures. To capture RPE, a "dark-field" mode may be employed; "off aperture" enhances light-scattering structures such as retinal ganglion cells (RGC) , and "split-detector" mode helps define the photoreceptor out segments. 
Imaging with AO-OCT resembles the technique used with AO-SLO. In fact, some AO-SLO machines combine SLO and OCT features to allow for both modalities to function in the same eye. In these machines, a scanning slow vertical mirror first captures a single SLO frame followed by a fast OCT frame in the vertical and horizontal planes, allowing for volumetric acquisition.  Within the broad category of AO-OCT modalities, several subdivisions fall under the technological descriptors of "time domain" and "spectral domain," which indicate the detection mechanism of the OCT signal.  AO-OCT is limited by motion artifact and fixation deficits of the patient, given its reliance on high-speed imaging and high magnification. Specifically, conditions including vitreous or aqueous opacity, age related related miosis, and loss of structural integrity to the retina may diminish the quality of imaging notably. 
The primary benefit of all of these imaging modalities is the ability of the user to visualize retinal structures in vivo on a cellular level. In terms of individual system benefits, they differ: AOFIO alone has been developed and approved for clinical use. In comparison to the scanning techniques of AO-OCT and AO-SLO, AO-FIO also offers a larger, and more rapidly acquired retinal image, reducing the procedure time. In contrast, AO-OCT notably allows for increased lateral resolution compared to other combined AO modalities and remains the most ideal for appreciating depth in the retina. Specifically, AO-OCT can aid in visualizing RGC, RPE, and choriocapillaris, which focus at varying depths. AO-SLO has demonstrated increasing utility in research as an excellent modality for visualizing cone photoreceptor inner segments, RGC, RPE, leukocyte migration in retinal vasculature, and single photoreceptors.  From a more clinical standpoint, AO-SLO offers insight into the physiology of conditions such as cone-packing density in healthy eyes, cone-packing density and vascular abnormalities such as microaneurysms in diabetic retinopathy, cone mosaic abnormalities in central serous chorioretinopathy (CSCR), early detection of drusen in eyes with AMD, and patterns of photoreceptor loss in inherited retinal diseases such as Stargardt and retinitis pigmentosa. 
Currently, all AO devices besides the rtx-1 AO-FIO device by Imagine Eyes are restricted to research use.  Implementation of AO in the clinical setting remains hindered by the cost and technical prowess required to run the system, the time investment of providers and patients needed to scan the eyes, the inconsistent quality of the images collected, and the lack of a standardized database to aid in interpretation of the images.  As a result, these challenges, among others, preclude the production of AO for commercial use at this time; yet, solutions are continually being pursued to overcome these difficulties in order to introduce AO into the clinical setting.
- Max, Claire. Introduction To Adaptive Optics And Its History. (2001).
- Burns SA, Elsner AE, Sapoznik KA, Warner RL, Gast TJ. Adaptive optics imaging of the human retina. Prog Retin Eye Res. 2019 Jan;68:1-30. doi: 10.1016/j.preteyeres.2018.08.002. Epub 2018 Aug 27. PMID: 30165239; PMCID: PMC6347528.
- Akyol, E., Hagag, A.M., Sivaprasad, S. et al. Adaptive optics: principles and applications in ophthalmology. Eye 35, 244–264 (2021). https://doi-org.proxy.cc.uic.edu/10.1038/s41433-020-01286-z
- Bedggood P, Metha A. Adaptive optics imaging of the retinal microvasculature. Clin Exp Optom. 2020 Jan;103(1):112-122. doi: 10.1111/cxo.12988. Epub 2019 Dec 3. PMID: 31797452.
- Porter, J. (2006). Adaptive optics for vision science : principles, practices, design, and applications . Wiley-Interscience. https://doi.org/10.1002/0471914878
- Imagine-Eyes. (n.d.). rtx-1 Adaptive Optics Retinal Camera. https://www.imagine-eyes.com/products/rtx1/#vid2
- Zawadzki, Robert & Capps, Arlie & Kim, Daeyu & Panorgias, Thanasis & Stevenson, Scott & Hamann, Bernd & Werner, John. (2014). Progress on Developing Adaptive Optics–Optical Coherence Tomography for In Vivo Retinal Imaging: Monitoring and Correction of Eye Motion Artifacts. Selected Topics in Quantum Electronics, IEEE Journal of. 20. 1-12. 10.1109/JSTQE.2013.2288302.
- Pircher, M., & Zawadzki, R. J. (2017). Review of adaptive optics OCT (AO-OCT): principles and applications for retinal imaging [Invited]. Biomedical optics express, 8(5), 2536–2562. https://doi.org/10.1364/BOE.8.002536