Abstract Precise tumor margin delineation remains critical for effective cancer treatment, yet traditional histopathological methods often miss early molecular changes indicative of malignancy. We developed artificial intelligence-enhanced mid-infrared photothermal imaging (AI-enhanced MIP), a label-free platform for automated spatial profiling of patient tissues at submicron resolution by acquiring intrinsic molecular spectroscopic information at each pixel. Analysis of surgical specimens from patients with head and neck squamous cell carcinoma at various pathological stages revealed spectroscopically defined transition zones (SDTZs) near tumor margins—regions that appear histologically normal but exhibit distinct metabolic alterations. Single-nucleus RNA sequencing validation confirmed elevated SREBF1 and CPT1A expression along with epithelial-mesenchymal transition signatures in SDTZs. Machine learning-assisted identification of SDTZ molecular signatures further improved staging prediction accuracy by 3.7-fold compared with using tumor core assessment alone. For clinical translation, AI-enhanced MIP requires no tissue pretreatment and enables rapid assessment using four selective wavenumbers, providing more precise tumor margin definition during intraoperative evaluation that could reduce recurrence while avoiding treatment delays.

Abstract Imaging scenes with a high dynamic range (HDR) of light intensities is critical for applications such as biomedical imaging, astronomical observation, and industrial automation, where accurate detection of both bright and dark regions is essential for precise analysis and decision-making. In this paper, we propose an HDR imaging approach harnessing optical forward scattering effect that breaks the limitations of image processing type. Our approach integrates a nonlinear deconvolution method based on speckle background noise estimation, along with Cross-correlation and Laplacian pyramid fusion method, to improve imaging precision and adaptability. By utilizing a digital micromirror device and a scattering diffuser, we develop a proof-of-concept experimental system, validating the effectiveness of reconstruction of faint details in HDR scenes. This method achieves dynamic range expansion from a 130.01 dB HDR scene using a detector with an 88.5 dB dynamic range, achieving a 119-fold intensity difference. Our work demonstrates a promising new solution for HDR imaging in demanding lighting environments, which could expand the scope of photoelectronic imaging application.

Abstract Three-photon microscopy (3PM) enables high-resolution three-dimensional (3D) imaging in deeply situated and highly scattering biological specimens, facilitating precise characterization of biological morphology and cellular-level physiology in vivo. However, the use of fluorescent probes with relatively low three-photon absorption cross-sections necessitates high-peak-power lasers for excitation, which poses inherent risks of light-induced damage. Additionally, the low repetition frequency of these lasers prolongs scanning time per pixel, hampering imaging speed and exacerbating the potential for photodamage. Such limitations hinder the application of 3PM in studying vulnerable tissues, including muscle regeneration. To address this critical issue, we developed the Multi-Scale Attention Denoising Network (MSAD-Net), a precise and versatile denoising network suitable for diverse structures and varying noise levels. Our network enables the use of lower excitation power (1/4–1/2 of the common power: 1.0–1.5 mW vs 4–6 mW) and shorter scanning time (1/6–1/4 of the common time: 2–3 μs/pixel vs 12 μs/pixel) in 3PM while preserving image quality and tissue integrity. It achieves a structural similarity index (SSIM) of with an average of 0.9932 and a fast inference time of just 80 ms per frame which ensured both high fidelity and practicality for downstream applications. By utilizing MSAD-Net-assisted imaging, we characterize the biological morphology and functionality of muscle regeneration processes through deep in vivo five-channel imaging under low excitation power and short scanning time, while maintaining a high signal-to-noise ratio (SNR) and excellent axial spatial resolution. Furthermore, we conducted high axial-resolution dynamic imaging of vascular microcirculation, macrophages, and ghost fibers. Our findings provide a deeper understanding of the mechanisms underlying muscle regeneration at the cellular and tissue levels.

Abstract Recent advancements in reconnaissance technologies necessitate the development of adaptive camouflage that effectively operates across multiple wavebands and scenarios. However, traditional adaptive camouflage technologies are predominantly limited to dynamic control of a single band in visible (VIS) or infrared (IR) band. In this study, we propose an integrated platform that acheives decoupled, dynamic control of visible and infrared signatures under extreme ambient temperatures. This device features a thermochromic (TCM) layer on top, a multi-walled carbon nanotube (MWCNT)-based emissivity-electrochromic tri-layer structure in the middle, and a thermoelectric device (TED) at the bottom. The IR-transparent TCM layer enables a color conversion from green to yellow at 28 °C, suitable for VIS camouflage in oasis and desert environments. Additionally, the device features an emissivity regulation from 0.44 to 0.84 (8–14 μm) and surface temperature adjustments between 10 °C and 60 °C, resulting in an impressive radiative temperature difference of 67.7 °C (-21.6 °C to 46.1 °C). The outstanding performance of our VIS-IR adaptive camouflage device illustrates its feasibility in extreme environments with significant diurnal temperature variations. The proposed device provides a new strategy in VIS-IR adaptive camouflage, paving the way for further advancements in camouflage device design and expanding its applications.

Abstract Solar concentrators laterally converge solar energy to the side of architectural glass and are attractive candidates for building-integrated photovoltaics. Present available luminescent-type and scattering-type solar concentrators suffer from omnidirectional waveguide induced low efficiency, coloring/hazing restrained aesthetic quality, and poor compatibility with existing architectural glass. Here, we propose a diffractive solar concentrator via directly coating cholesteric liquid crystal (CLC) layers onto the architectural glass. The stacked CLC layers with submicron lateral periodic alignment enable broadband and unidirectional waveguiding inside the glass, and thus supply a high-efficient platform for transmissive solar energy capturing with merits of high aesthetic quality and economic viability. A 1-inch-diameter prototype powers a 10-mW fan outdoors, and a typical 2-m-wide window is calculated to concentrate solar energy by 50 times. The design is expected to bring a global terawatt-scale green energy supply and billion-ton annual carbon emission reduction, meeting with the sustainable development of human society.

In response to the growing demands of advanced 5G/6G communication technologies, millimeter-wave vortex beams have emerged as a promising solution to increase channel capacities. This paper introduces a novel and efficient method for vortex beam generation by leveraging the intrinsic singularities of dipole scatterers and enhancing their performance through non-local coupling. We demonstrate that the intrinsic singularities—amplitude-zero points in the scattering patterns of electric dipole (ED) and magnetic dipole (MD) resonances – enable the conversion of spin angular momentum (SAM) into orbital angular momentum (OAM), generating a vortex electric field distribution. By arranging these dipolar units into a periodic array, we establish a dual-resonance non-local metasurface that improves directivity and efficiency via non-local collective interactions and the generalized Kerker effect. This configuration significantly enhances forward scattering, producing highly directional vortex beams. Our experimental results show that the non-local metasurface achieves a vortex conversion efficiency approximately 2.2 times higher than that of a reference structure around 40 GHz. This alignment-free, high-efficiency solution offers great potential for expanding millimeter-wave communication capacity and advancing photonic applications.

Holographic 3D display technology, widely considered the ultimate solution for real 3D display, has broad applications in fields including advertisement, industrial manufacturing and military. However, it is difficult to simultaneously realize color holographic 3D display with wide viewing angle and high brightness required for an immersive visual experience. Here, a novel holographic 3D display system based on a customized achromatic liquid crystal grating and a phase-only spatial light modulator is proposed. Thanks to the secondary diffraction performed by the achromatic liquid crystal grating, nine secondary diffraction images of red, green and blue channels overlap in space in time sequence. Additionally, a high brightness hologram encoding method is developed, which introduces a frequency loss function with dynamic weights to ensure that differences of all frequency components in the frequency domain can be learned. The proposed method dramatically enhances light energy utilization by a factor of five, resulting in significantly brighter reconstructed images while substantially attenuating background noise in non-target regions. This groundbreaking system, achieving a remarkable ~ 65° wide viewing angle with good image quality and high brightness, represents a significant advancement in holographic technology, offering a comprehensive solution for wide-viewing-angle, high-brightness, color 3D displays with potential applications across diverse technological domains.

Soliton molecules (SMs) play a crucial role in nonlinear optical systems, enriching our understanding of nonlinear science through the study of their interaction dynamics. While passively mode-locked fiber lasers offer an efficient platform for generating diverse types of SMs, the complex internal dynamics of the laser often pose challenges in achieving predetermined temporal separations between SMs. Here, we implement a delayed optical feedback technique within a femtosecond optical parametric oscillator, enabling the generation of SMs with precise and controllable temporal separations. A theoretical model, which models the intracavity iterations of the signal with a simplified Ikeda map, is proposed to study the impact of parametric gain, intracavity feedback delay, and cavity length on the internal separations of the SMs. Our experimental results confirm that adjusting the cavity length allows for producing desired temporal separations within SMs. To reveal the evolution dynamics of the SMs, we further develop a rigorous numerical model using the carrier-resolved Forward Maxwell Equation, which is capable of modeling ultra-broadband complex dynamics based on a single equation without relying on the slowly-varying envelope approximation. The numerical model unveils the rich formation dynamics of the SMs at various separations, which confirms the critical role of the gain window provided by the pump. This work opens up new opportunities for the on-demand generation of SMs and provides valuable insights into the complex dynamics in femtosecond optical parametric oscillator systems with optical delayed feedback.

Information security plays an important role in every aspect of life to protect data from stealing and deciphering. However, most of the previously reported works were based on pure algorithm layer or pure physical layer encryptions, which have certain limitations in security. In this paper, a nondeterministic message encryption communication scheme is proposed based on a spin-space-frequency multiplexing metasurface (SSFMM), which integrates both algorithmic and physical layer encryptions, and can also produce multiple different ciphertexts for the same message to prevent the message from being cracked through frequency analysis, thus greatly enhancing the security of the information. To be specific, an SSFMM is first designed as a physical-layer meta-key, which can generate eight independent dot matrix holograms with different spin, space, and frequency characteristics. The target message is then encrypted based on these dot matrix holograms combined with algorithmic operations, and the encrypted message is converted into a quick response (QR) code for easy sending to the target users. Once the target user gets that QR code, he/she can scan it to obtain the encryption information, and then recover the target message according to the pre-agreed encryption protocol combined with the eight dot matrix holograms of SSFMM. Finally, the feasibility of the proposed encryption scheme was experimentally validated at the microwave frequency band.