Research Areas

In vivo optical microscopy aims to observe (sub)cellular structures and functional dynamics within living biological tissues or model organisms in situ and noninvasively. Compared with thin samples such as tissue sections or adherent cultured cells, three-dimensional in vivo optical microscopy has two defining characteristics—which also represent major technical challenges.

First, it requires optical sectioning (depth-resolved imaging) capability to enable three-dimensional imaging of thick biological samples without physical sectioning, thereby allowing in situ imaging of intact living tissue.
Second, it requires sufficiently high four-dimensional (3D space × time) spatiotemporal resolution to capture high-throughput structural and functional dynamics in living samples or organisms, while effectively compensating for motion artifacts caused by physiological activity.

Advances in basic life sciences—particularly neuroscience—and precision medicine continue to raise new challenges for in vivo optical microscopy. Two representative examples are outlined below.

• Understanding brain function requires monitoring neural activity across the entire brain—or even the whole body—of freely behaving small model organisms such as C. elegans, Drosophila (larvae), and zebrafish at single-neuron resolution. However, commonly used fluorescence microscopy techniques with optical sectioning capabilities remain limited in volumetric imaging speed, phototoxicity control, and sample accessibility, making them insufficient for long-term, large-scale observation of neural population dynamics across the brain.

• In clinical medicine, traditional biopsy followed by histopathological analysis, although considered the current gold standard for diagnosis, is inherently invasive, ex vivo, and non–real-time. This approach loses the intrinsic three-dimensional structure and dynamic functional information of living tissues and still carries risks of sampling error and delayed diagnosis. Developing high-throughput, in situ three-dimensional pathological imaging methods that achieve “histology-like imaging without tissue sectioning” represents an important direction for future precision diagnostics. Such techniques could provide real-time pathological information while minimizing tissue damage, improving diagnostic efficiency, and ultimately enhancing patient outcomes.

To address these challenges, our research group integrates the academic and technological resources of the University of Science and Technology of China (USTC) and actively promotes interdisciplinary collaborations across science, engineering, biology, and medicine.

On the one hand, we develop next-generation SCAPE (Swept Confocally-Aligned Planar Excitation) microscopy methods and instrumentation, aiming to substantially improve key performance metrics such as spatiotemporal resolution, fluorescence collection efficiency, and field of view, thereby enabling long-term observation of whole-brain neural activity in freely behaving model organisms.

On the other hand, we design customized handheld or endoscopic miniature volumetric imaging probes tailored to the needs of different clinical specialties. These systems incorporate innovative optical architectures to enable high-resolution, minimally invasive or noninvasive in situ three-dimensional pathological imaging, supporting applications such as early disease screening, optical biopsy assistance, and microsurgical navigation, and facilitating translational research at the interface of medicine and engineering.

Selected Previous Research Contributions

1. Compact SCAPE microscope for real-time in vivo volumetric histopathology imaging

Developed the first miniaturized SCAPE microscope designed for real-time in vivo volumetric pathology imaging in clinical settings. Despite reducing the main objective aperture diameter by two-thirds and shrinking the system footprint by four-fifths, the compact SCAPE microscope achieved fluorescence collection efficiency and three-dimensional resolution (0.81 µm × 1.07 µm × 2.10 µm) comparable to conventional benchtop counterparts. This enabled large-field, panoramic volumetric imaging in vivo using purely endogenous autofluorescence.

The high spatiotemporal resolution of SCAPE microscopy allows effective compensation for tissue motion, making it a promising approach for “histology-like imaging without sectioning” across large tissue volumes. For clinical applications, however, a key requirement is the development of compact or miniature SCAPE probes. Conventional benchtop SCAPE microscopes rely on highly complex optical architectures that are difficult to miniaturize.

Starting from the fundamental principles of SCAPE microscopy, I first integrated the light-sheet generation optics and fluorescence detection objective into a rear orthogonal illumination–detection module, eliminating the dichroic beam splitter and branching optical paths in traditional designs. This significantly simplified the optical layout. I then designed a front-end cascaded 4f optical relay bridging the secondary objective and the primary objective, conjugating the virtual sample space of the rear module with the open sample space of the primary objective. Through optmized optical folding strategies, the front-end module was made more compact.

Animal experiments demonstrated that this compact SCAPE microscope could clearly visualize cellular structures in mouse organs such as liver and kidney, using only endogenous autofluorescence. In imaging experiments on the unlabeled tongue of healthy volunteers, volumetric imaging rates of 4.8 volumes per second were achieved with only 6.4 mW excitation power. Even when the subject moved the tongue freely, the system could estimate the three-dimensional displacement between consecutive data blocks, thanks to the high-speed volumetric rate, and fuse the registered volumes into a multi-millimeter-scale panoramic 3D mosaic. These results demonstrate the system’s strong robustness against tissue motion and its potential for clinical applications.

2. Fiber-scanning miniature two-photon endomicroscopy

Led the development of two generations of flexible fiber-scanning miniature two-photon endomicroscopes.

2.1 First-generation miniature two-photon endomicroscope

Conducted a systematic investigation of the factors limiting the signal-to-noise ratio (SNR) of fiber-based two-photon endomicroscopy and developed several key solutions:

1. Identified a previously overlooked source of background noise—nonlinear luminescence generated in germanium-doped fiber cores under pulsed excitation—and reduced this noise to less than one-quarter of that in commercial double-clad fibers using a custom double-clad fiber with an undoped pure-silica core.

2. Established a backward fluorescence collection model for two-photon endomicroscopy, which guided the optimization of chromatic aberration compensation in custom miniature GRIN objectives and the optical expansion of the inner cladding of custom double-clad fibers, increasing fluorescence collection efficiency by approximately 2.5×.

3. Developed a nonlinear time–frequency joint dispersion compensation method that mitigates spectral compression caused by self-phase modulation, improving two-photon excitation efficiency by 2–3× compared with conventional linear compensation methods.

Based on these innovations, I built a first-generation miniature two-photon endomicroscope probe with a diameter of only 2.2 mm, achieving more than 20× improvement in detection sensitivity compared with previous designs and overcoming the long-standing SNR limitations of fiber-based two-photon endomicroscopy.

2.2 Second-generation two-photon endomicroscope with cascaded NA amplification and much improved space–bandwidth–speed product 

In fiber-cantilever resonant scanners, one typically needs to trade-off between the field of view (proportional to cantilever length) and scan speed (proportional to resonant frequency). I theoretically demonstrated that the space–bandwidth–speed product of a resonant fiber-scanning endomicroscopes is proportional to the square of the output NA of the cantilever. Based on this insight, I further proposed a cascaded numerical aperture (NA) amplification strategy and constructed a single-mode composite fiber cantilever with an effective NA of ~0.35, breaking the conventional NA limit (~0.12) of standard single-mode fibers.

Using this design, I developed second-generation two-photon endomicroscopes with significantly improved imaging throughput. In specific, while maintaining the same spatial resolution and excitation efficiency, the 2nd-gen probes could achieve either: (1) 3× larger field of view at the same imaging speed (≈9× increase in space–bandwidth product), or (2) 3× higher imaging frame rate with a 1.5× larger field of view. These results validated both the feasibility and flexibility of the cascaded NA amplification strategy.

2.3 Label-free metabolic imaging with two-photon endomicroscopy

I further developed a label-free two-photon endomicroscopy method capable of simultaneously measuring the optical redox ratio and NADH fluorescence lifetime, enabling the first demonstration of simultaneous structural and metabolic imaging in vivo using two-photon endomicroscopy.

By analyzing the photon arrival time distribution of background signals and the weak photon flux of NADH fluorescence, I developed a maximum-likelihood-estimation-based bi-exponential fitting algorithm that significantly improved the accuracy of fluorescence lifetime analysis, enabling quantitative differentiation between free and protein-bound NADH. Animal experiments demonstrated that this system can monitor dynamic changes in redox ratio and intracellular NADH states in mouse kidney proximal tubule epithelial cells during ischemia–reperfusion, as well as changes in NADH abundance and fluorescence lifetime during apoptosis in tumor cells in vivo.

Selected Publications

1. Zixian Cao, Jiapeng Zhu, Cheng Zhang, Qianqian Wang, Yankan Huang, Wei Liu, Bingxin Shen, Yuming Chai, Zhaomin Zhong, Li He, Quan Wen, Han Wang, and Wenxuan Liang†. Mesoscopic SCAPE Microscope with a Rescanned,
   Super-oblique Illumination Plane. bioRxiv 2025.04.25.650616; doi: https://doi.org/10.1101/2025.04.25.650616

2. Bingxin Shen, Zixian Cao, Yankan Huang, Jiapeng Zhu, Haiyan Huang, Wenxuan Liang†; Computationally designing a dual-aspheric compound lens toward multi-depth endomicroscopy imaging. Applied Physics Letters 126 (22), 223701
   (2025). https://doi.org/10.1063/5.0268424

3. Liang W†, Liu Y, Guan H, Sakulsaengprapha V, Luby-Phelps K, Mahendroo M, and Li X. Cervical Collagen Network Porosity Assessed by SHG Endomicroscopy Distinguishes Preterm and Normal Pregnancy — a Pilot Study. 
   IEEE Transactions on Biomedical Engineering, 72(2), 777-785 (2025) †Correspondence    DOI: 10.1109/TBME.2024.3472015
   

4. Hall G*, Liang W*†, Bhujwalla ZM, and Li X†. SHG fiberscopy assessment of collagen morphology and its potential for breast cancer optical histology. 
   IEEE Transactions on Biomedical Engineering 71(8), 2414-2420 (2024). *Equal contribution, †Correspondence     DOI: 10.1109/TBME.2024.3372629
   

5. Liang W, Chen D, Guan H, Park HC, Li K, Li A, Li MJ and Li X. Label-Free metabolic imaging in vivo by two-photon fluorescence lifetime endomicroscopy. 
   ACS Photonics, 9(12), 4017-4029 (2022)         https://doi.org/10.1021/acsphotonics.2c01493

6. Patel KB, Liang W, Casper MJ, Voleti V, Zhao HT, Perez-Campos C, Liu JM, Coley SM, and Hillman EMC. High-speed light-sheet microscopy for the in-situ acquisition of volumetric histological images of living tissue. 
   Nature Biomedical Engineering 6, 569-583 (2022)         https://doi.org/10.1038/s41551-022-00849-7

7. Liang W, Park HC, Li K, Li A, Chen D, Guan H, Yue Y, Gau YT, Bergles DE, Li MJ, Lu H, and Li X. Throughput-speed product augmentation for scanning fiber-optic two-photon endomicroscopy. 
   IEEE Transactions on Medical Imaging 39(12), 3779-3787 (2020)         https://doi.org/10.1109/TMI.2020.3005067

8. Li K*, Liang W*, Yang Z, Liang Y, and Wan S. Robust, accurate depth-resolved attenuation characterization in optical coherence tomography. 
   Biomedical Optics Express 11(2), 672-687 (2020) *Equal contribution         https://doi.org/10.1364/BOE.382493

9. Voleti V, Patel KB, Li W, Campos CP, Bharadwaj S, Yu H, Ford C, Casper MJ, Yan RW, Liang W, Wen C, Kimura KD, Targoff KL, and Hillman EMC. Real-time volumetric microscopy of in vivo dynamics and large-scale samples with SCAPE 2.0. 
   Nature Methods 16(10), 1054–1062 (2019)         https://doi.org/10.1038/s41592-019-0579-4

10. Li K, Liang W, Mavadia-Shukla J, Park HC, Li D, Yuan W, Wan S, and Li X. Super-achromatic optical coherence tomography capsule for ultrahigh‐resolution imaging of esophagus. 
    Journal of Biophotonics 12(3), e201800205 (2019)         https://doi.org/10.1002/jbio.201800205

11. Liang W*, Hall G*, and Li X. Spectro-temporal dispersion management of femtosecond pulses for fiber-optic two-photon endomicroscopy. 
    Optics Express 26(18), 22877-22893 (2018) *Equal contribution         https://doi.org/10.1364/OE.26.022877

12. Liang W, Hall G, Messerschmidt B, Li MJ, and Li X. Nonlinear optical endomicroscopy for label-free functional histology in vivo. 
    Light: Science and Applications 6, e17082 (2017)         https://doi.org/10.1038/lsa.2017.82