In confocal microscopy the 3D resolution is mainly achieved from the light originating from focus and not from the scattered light, a common problem associated mainly while imaging deep tissues [1–3]. Light scattering could be reduced using longer excitation wavelengths [4–6], however, it is limited to certain wavelengths. Linear (one-photon) excitation microscopy is mainly suited for near-surface (<100 μm) imaging. While non-linear (two-photon fluorescence) microscopy provides the advantage of imaging brain regions >100 μm with minimal scattering and better resolution. Deep tissue imaging at the single-cell level has provided adequate information on synaptic activity and function. Table 1 compares the with and without adaptive optics based confocal and two photon fluorescence microscopy. Conventional functional visualization neuroimaging techniques such as magnetic resonance imaging (MRI), positron emission tomography (PET), computerized tomography (CT) is currently used for imaging the human brain in vivo are usually expensive and have a poor spatial and temporal resolution [7]. The improvements in the optical approach have developed techniques that allow non-labeled imaging of tissues with high sensitivity, specificity, and spatial resolution under in vivo and in vitro conditions. Non-linear microscopy, in particular, TPF supplemented by the recent advances in optical labeling of deep tissues has become the method of choice for imaging intact tissues and live animals [4, 5, 8]. Principally, the TPF technique utilizes near-infrared wavelength and ultrashort pulses where the interaction of two photons usually of higher-order is responsible for contrast generation [9, 10]. Due to its high fluorescence intensity and depth resolution [2], TPF has drawn substantial attention for the study of deeper structures and vital functions of the brain and eye. Indeed, TPF coupled with fluorescence lifetime imaging (FLIM) has shown promising data on vitamin A metabolism in the eye to study the early onset of retinal diseases [11]. Owing to the heterogeneity of biological samples under investigation and the optical setup used, wavefront distortions limit the fluorescence signal-to-noise ratio and resolution with increasing depth [12]. AO based TPF overcomes such a loss in resolution by detecting and correcting the type of wavefront aberration generated while imaging scattering samples [13]. In the following sections, we will discuss some of the applications of AO in TPF based brain imaging.

Over the past decades, there have been great improvements in the field of ophthalmoscopy which has not only allowed scientists to successfully resolve fine retinal structures but also improved the clinical interventions for different eye diseases. This was mainly possible from the development of optical systems and recent advances in aberration correcting systems. However, when augmented with existing ophthalmoscopes such as conventional fundus imaging, SLO and spectral domain-optical coherence tomography (SD-OCT), AO enhances both structural and functional retinal imaging. First developed by David Williams and coworkers at the University of Rochester, AO-fundus imaging was used to overcome wavefront aberrations in <0.5 s and to obtain dynamic correction of fluctuations in Zernike mode [14]. SLO was first applied in the 1980s and is known to be the method of choice for clinical imaging with superior contrast and resolution. In SLO, retinal images are generated over time by measuring the scattered light as the focused beam scans across the retina surface. SLO provides enhanced image contrast mainly due to the “confocality” wherein light not originating from focus is rejected via a pinhole. Reflectance imaging is an application of SLO where the scattered light within an optically aberrated biological sample like an eye is measured. Such a technique, when used along with different wavelength lasers, could be used in clinical settings for multicolour imaging like fundus autofluorescence (FAF) [15]. Although, being routinely used in clinical imaging of the eye, contrast and depth of imaging obtained in conventional SLO methods are limited. In the later part of the review, we highlight some of the recent advances in AO based retinal imaging.

Optical imaging is the foremost important method of choice for imaging biological specimens. With the recent developments in optical methods, three-dimensional and highly resolved structures from biological samples can be achieved. One of the limitations of deep brain imaging is the scattering of light through the heterogeneous tissues, which affects the penetration depth and image resolution. The advances in the TPF microscopy techniques provide a powerful tool for overcoming scattering of light, for in vivo neuro-imaging with sub-cellular resolution, endogenous contrast specificity, pinhole less optical sectioning capability, high penetration depth, and so on. The NIR light source is usually the choice for excitation source as it avoids water absorption, thus, reducing photo-bleaching and photo-toxicity in the biological optical window, and also increases the penetration depth within the thick tissue [2, 3, 16]. Ability to image deeper brain regions with the single-cell resolution has refined the establishment of synaptic activity and its functions. The high fluorescence intensity and depth resolution assist in studying the structure and vital functions of the brain.

An optical system incorporates three major components, the laser system (pulsed), optical modalities (lenses, mirrors and objective etc.) and the specimen under investigation. Each of these components contributes toward the optical aberrations which in turn affects the imaging time and resolution. Although using aberration-corrected optical components reduce the imaging limitations when scanning deeper regions of a biological specimen, sample induced aberrations and scattering yet remains substantial. One such tool very quickly realized and implemented in optical imaging of biological samples is AO. For thick biological specimens, the optical signal strength reduces due to the aberration caused by the scattering medium (sample) and optical components used in the system [17, 18]. The ability to correct the aberration and improve the signal strength has been demonstrated using adaptive optics integrated into a wavefront sensor (WS) or sensor-less AO [19]. In the case of sensor-based AO, a WS estimates the aberration present in the imaging system and either a liquid crystal spatial light modulator (LC-SLM) or deformable membrane mirror membrane (DMM) corrects the aberration. Whereas, in sensor-less AO, the wavefront of the excitation beam is modulated and the optimal wavefront is determined through quantifying the properties of the acquired image in the form of intensity or sharpness in a feedback loop [18]. Sensorless corrections have particularly suffered from long optimization time and the introduction of fluorescent fiducial markers. In the following sections, we will briefly discuss the applications of AO in biological imaging using TPF and SLO.

The introduction of AO-based optical imaging has greatly improved deep-tissue imaging which was previously unattainable using conventional optical imaging microscopes. Since its usage in correcting atmospheric aberrations in telescopes, AO-based optical imaging has now paved its way into biological imaging. With AO, the diffraction-limited resolution could be easily achieved by correcting deep-tissue mediated aberrations. More importantly, the technology has improved the imaging quality and speed at which fine structural and functional aspects of living biological samples can be studied. The direct wavefront corrections (using SH wavefront sensors) provides a rapid and effective imaging modality compared to indirect wavefront sensing methods. Although the latter could be easily applied to any existing microscopes and is suitable for both scattering and transparent samples. Imaging deeper regions within a biological tissue have mostly been overcome by the implementation of AO based corrections. Although imaging complex neuronal populations with an AO-based optical system needs further improvements, there has been considerable improvement in retinal imaging irrespective of the differences arising from refractive index or age. Yet, imaging deeper neuronal populations with enhanced spatial-temporal resolution in a freely moving animal remain challenging. While recording neural activity in awake animals, brain movement remains a serious concern and are corrected by post image processing of 2D images. However, this approach is slow and cannot correct axial movements in a small region of interest and photostimulation. In 2016 Nadella and coworkers developed a 3D random-access two-photon laser scanning microscope using a rapid closed-loop system and acousto-optic lens (AOL). This system improves motion-mediated artifacts and monitors real-time 3D brain tissue movement at up to 1 kHz with submicrometer spatial resolution in behaving mice and zebrafish [47]. In another study, a compact AOL two-photon microscope was used for high-speed line scanning within an imaging volume at 35–50 kHz up to hundreds of micrometers. It was able to monitor cerebellar interneurons sparsely distributed neuronal activity over large depth ranges in awake animals [48].

Miniaturized fluorescence microscopes or miniscopes are becoming a key tool in neuroscience research owing to their high-resolution deep brain imaging in freely behaving animals. Owing to their lightweight, miniscopes utilizing one-photon or two-photon excitation modalities have been reported in the study of cellular activity when coupled to a freely moving animal [49, 50]. Miniscopes provides convenient in vivo deep-tissue imaging in an awake free moving animal due to, i) it's decreased footprint and weight ii) easy mounting on small animals iii) ability to record multiple neuronal activities simultaneously, thereby aiding in better understanding inter-regional signaling in behaving animal models [51].

Two-photon microscopy serves as a method of choice over conventional microscopy techniques for deep-tissue imaging. However, due to the scattering of ballistic photons from deeper samples, an optical method that could limit photon loss is necessary. TPF augmented with AO provides the advantage of limiting photon loss from deep biological samples such as the human brain [3] and eye [35]. AO-TPF has been incorporated in neurobiology and ophthalmology labs across the globe to achieve diffraction-limited contrast in biological imaging [14, 52, 53]. The use of AO-based optical imaging in ophthalmology has greatly changed the way we see the human eye both at the cellular and metabolic level. Integrating sensorless AO along with dispersion compensation of ultrashort laser pulse can improve TPF intensity of endogenous retinal fluorophores. Indeed, the use of ultrashort laser (20-fs pulse) did not show any modifications in function as well as the structure of the retina in vivo. Interestingly, the broad bandwidth of short pulses excites multiple fluorophores which provides better applications while studying cellular metabolism [11]. AO-based ophthalmoscopes also provide a platform to study several genetic diseases of the eye. AO-SLO has not only improved eye imaging but also facilitated its application in clinics for studying and monitoring retina-related conditions such as macular degeneration and Stargardt disease [54]. Above all, with the current advances in the filed of AO and TPF scientists can effortlessly image different regions within the brain or eye of a live organism which was impossible previously. With further improvements in the existing fluorescent probes, AO-based biological imaging modalities would further improve our way of seeing things in vivo.

PS wrote the manuscript. NM structure the manuscript and wrote the manuscript. Both authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.