Keynote Speakers

Speech Title: A Tale of Force and Light: How Optical Coherence Tomography and Raman Spectroscopy, Combined with Mechanical Testing, Reveal the Dynamic Behavior of Healthy and Diseased Articular Cartilage

Abstract: Articular cartilage achieves long-term load bearing and low-friction joint motion through a finely organised, depth-dependent collagen architecture and tightly regulated biochemical environment. Early disruption of this organisation alters tissue mechanics and molecular behaviour well before gross structural damage, such as osteoarthritis (OA), becomes apparent. However, conventional approaches rely on destructive, end-point analyses that preclude observation of cartilage dynamics under load. Here, we present a multimodal, non-destructive platform combining polarisation-sensitive optical coherence tomography (PS-OCT), Raman spectroscopy, and controlled mechanical compression to interrogate the coupled mechanical, structural, and biochemical response of cartilage in real time.
We have validated a compression–imaging platform, which enables simultaneous acquisition of depth-resolved PS-OCT images, Raman spectra, and force–displacement data during in situ controlled loading. Using bovine cartilage-on-bone samples, the system reproducibly captures classical microanatomical deformation patterns, including chevron-shaped shear deformation and formation of a shear discontinuity, while revealing their temporal evolution during loading. Dynamic birefringence and Raman measurements demonstrate clear distinctions between healthy and artificially damaged cartilage within the first minutes of compression, highlighting the sensitivity of the approach to early mechanical dysfunction.
This platform has been applied to study a range of cartilage tissue, from healthy to those with their surface removed, and cartilage with clear signs of degeneration. The aim is to elucidate mechanisms underlying the initiation of OA. PS-OCT reveals that healthy tissue exhibits a gradual, depth-dependent increase in birefringence with strain that stabilises over time, consistent with controlled collagen reorientation and effective load redistribution. In contrast, damaged and diseased tissues show rapid, irregular birefringence changes, loss of coherent shear boundaries, and reduced lateral load dissipation. Complementary Raman spectroscopy, analysed using two-dimensional correlation spectroscopy, shows coordinated protein-, proteoglycan-, and water-related biochemical responses in healthy cartilage, while diseased tissue exhibits faster, altered biochemical dynamics and disrupted molecular coupling.
Together, these results demonstrate that healthy cartilage uniquely maintains coordinated mechanical, structural, and biochemical dynamics under load. The multimodal platform provides sensitive optical biomarkers of early degeneration and offers a powerful tool for early OA detection, functional tissue assessment, and evaluation of regenerative and therapeutic strategies.

Speech Title: Advances in Tm-doped Fibers for Power Scaling of 2 µm and 0.82 µm Lasers and Amplifier

Abstract: Thulium (Tm-) doped fiber lasers (TDFLs) operating in the eye-safe 2 μm spectral region have demonstrated significant power scaling capability, following Yb-doped fiber lasers operating at 1 μm. Their high quantum efficiency (>1.9) under ~790nm laser diode pumping, enabled by a “two-for-one” cross-relaxation process, allows efficient high-power fibre laser operation, with demonstrated output powers of around 1 kW. However, the high thermal load in the fiber (on the order of several 100 W/m) presents significant thermal management challenges and can lead to fiber coating degradation and failure during high-power operation. Consequently, these thermal effects remain the primary limiting factor to further power scaling of 790 nm cladding-pumped TDFL systems beyond the kilowatt regime.
This presentation reviews progress in cladding-pumped Tm-doped silica fibers with optimized core compositions and engineered refractive index profiles for achieving high laser efficiency at 2 µm and to support high-power operation at shorter wavelength region (~0.82 µm), which is particularly challenging to achieve. It further discusses high-temperature resistant coating that help prevent degradation under high-power operation and supports robust and reliable laser operation.

Speech Title: A Photonic, Ultra-fast, and Resource-Efficient Neural Network (PURE-NN)

Abstract: Photonic neural networks (PNNs) show immense potential as a next-generation AI hardware paradigm. Yet, despite rapid hardware innovations, the gap between current PNN systems and practical AI applications remains significant; while effective as standalone components, PNNs are difficult to integrate into modern AI pipelines. To bridge this gap, this work proposes a versatile, digital micro-mirror device (DMD)-based PNN computing unit designed to minimize system complexity and electro-optic conversions. Furthermore, a GPT-inspired image task scheduler is introduced, which drives the PNN as a black-box processor to execute complex tasks.

Speech Title: Imaging Spheroid Morphology and Function by Computationally Augmented Optical Coherence Tomography

Abstract: Spheroids are essential samples for cancer research and regenerative medicine. While recent advancements in cultivation technology have enabled the creation of larger, more complex spheroids that better mimic human physiology, standard imaging modalities cannot adequately assess the three-dimensional morphology and functions of living spheroids. Here, we demonstrate a new label-free, three-dimensional imaging modality: computationally augmented optical coherence tomography (CA-OCT). This modality does not use exogenous agents to highlight spheroid functions; instead, it leverages intrinsic intracellular dynamics to generate three-dimensional functional images. Additionally, several deep-learning models are incorporated to enhance both the imaging speed and quality of CA-OCT. In this talk, the concept and technical details of CA-OCT, as well as its application to spheroid imaging, will be presented.

Speech Title: Atmospheric and Oceanic Lidar for Environmental Studies

Abstract: The complex and variable atmospheric-oceanic environment is of great significance to ecological protection, climate and meteorology, national defense, and military industries. Light Detection and Ranging (Lidar) possesses the advantages of cross-medium detection and high spatiotemporal resolution. When integrated with multi-platform Lidar systems, it enables large-scale, long-term continuous, and high-precision detection of atmospheric-oceanic parameters, exhibiting technical characteristics of cross-scale spatiotemporal fusion, multi-parameter high-precision coordination, and data closed-loop enhancement. Focusing on the "three-dimensional distribution of optical and microphysical properties of atmospheric aerosols, clouds, and marine particles in coastal areas," extensive experiments have been conducted on atmospheric-oceanic environmental Lidar systems with different platforms and mechanisms in China's inland lakes as well as offshore areas such as the East China Sea and the South China Sea, yielding a large amount of three-dimensional detection data of coastal ecological environment parameters in China.

Speech Title: Evidential Machine Learning: Theory and Applications to Medical Image Processing

Abstract: Evidential machine learning refers to a family of machine learning methods that quantify uncertainty using evidence theory, a generalisation of Bayesian theory based on belief functions. We first give a short introduction to belief functions and their use in information fusion and decision-making under uncertainty. We then describe machine learning models for classification and clustering grounded in evidence theory. Finally, we present applications to medical image processing, including cancer treatment outcome prediction and tumour segmentation.

Speech Title: Mueller Matrix Polarimetry of Metasurfaces

Abstract: Besides exhibiting significantly enhanced chirality, chiral nanostructures can serve as highly sensitive platforms for molecular enantiomer detection, offer control over multiple photonic degrees of freedom, permitting applications in optical manipulation, holographic encryption, structured illumination microscopy, chiral lasing, and quantum communication. Despite these advances, probing intrinsic chiral responses at the nanoscale remains a fundamental challenge in chiral nanophotonics, where multiple polarization effects coexist and often gets intertwined in complex ways. Here, we introduce a Mueller matrix–based methodology that captures the complete polarization response of complex nanostructured systems, enabling the decoupling of concurrent anisotropic effects and the accurate quantification of intrinsic chirality. We demonstrate that conventional approaches, such as circular dichroism, optical rotatory dispersion, and standard ellipsometric measurements can lead to significant artifacts when applied to systems exhibiting coupled polarization phenomena. In contrast, Mueller matrix polarimetry provides a robust framework to disentangle these effects and retrieve the true chiro-optical response. By integrating Mueller polarimetry with Fourier-domain imaging, we further resolve spatially inhomogeneous scattered polarization states, and establish a route toward advanced chiral sensing platforms. Beyond improved measurement accuracy, the proposed approach reveals deeper physical insights into intrinsic and extrinsic chirality in metasurfaces, polarization conversion mechanisms, and enhanced chiral light–matter interactions, which highlights that chiral nanophotonics encompasses a far richer landscape than that accessible through conventional chiro-optical measurements, opening new directions for both fundamental studies and practical applications.