Search

Abstract Multimodal deep learning plays a pivotal role in supporting the processing and learning of diverse data types within the realm of artificial intelligence generated content (AIGC). However, most photonic neuromorphic processors for deep learning can only handle a single data modality (either vision or audio) due to the lack of abundant parameter training in optical domain. Here, we propose and demonstrate a trainable diffractive optical neural network (TDONN) chip based on on-chip diffractive optics with massive tunable elements to address these constraints. The TDONN chip includes one input layer, five hidden layers, and one output layer, and only one forward propagation is required to obtain the inference results without frequent optical-electrical conversion. The customized stochastic gradient descent algorithm and the drop-out mechanism are developed for photonic neurons to realize in situ training and fast convergence in the optical domain. The TDONN chip achieves a potential throughput of 217.6 tera-operations per second (TOPS) with high computing density (447.7 TOPS/mm2), high system-level energy efficiency (7.28 TOPS/W), and low optical latency (30.2 ps). The TDONN chip has successfully implemented four-class classification in different modalities (vision, audio, and touch) and achieve 85.7% accuracy on multimodal test sets. Our work opens up a new avenue for multimodal deep learning with integrated photonic processors, providing a potential solution for low-power AI large models using photonic technology.

The focus of the research is on the label-constrained time-varying shortest route query problem on time-varying communication networks. To the best of our knowledge, research on this issue is still relatively limited, and similar studies have the drawbacks of low solution accuracy and slow computational speed. In this study, a wave delay neural network (WDNN) framework and corresponding algorithms is proposed to effectively solve the label-constrained time-varying shortest routing query problem. This framework accurately simulates the time-varying characteristics of the network without any training requirements. WDNN adopts a new type of wave neuron, which is independently designed and all neurons are parallelly computed on WDNN. This algorithm determines the shortest route based on the waves received by the destination neuron (node). Furthermore, the time complexity and correctness of the proposed algorithm were analyzed in detail in this study, and the performance of the algorithm was analyzed in depth by comparing it with existing algorithms on randomly generated and real networks. The research results indicate that the proposed algorithm outperforms current existing algorithms in terms of response speed and computational accuracy.

Matrix inversion is a fundamental and widely utilized linear algebraic operation but computationally expensive in digital-clock-based platforms. Optical computing is a new computing paradigm with high speed and energy efficiency, and the computation can be realized through light propagation. However, there is a scarcity of experimentally implemented matrix inverters that exhibit both high integration density and the capability to perform complex-valued operations in existing optical systems. For the first time, we experimentally demonstrated an iterative all-optical chip-scale processor to perform the computation of complex-valued matrix inversion using the Richardson method. Our chip-scale processor achieves an iteration speed of 10 GHz, which can facilitate ultra-fast matrix inversion with the assistance of high-speed Mach–Zehnder interferometer modulators. The convergence can be attained within 20 iterations, yielding an accuracy of 90%. The proposed chip-scale all-optical complex-valued matrix inverter represents a distinctive innovation in the field of all-optical recursive systems, offering significant potential for solving computationally intensive mathematical problems.

Abstract The increasing amount of data exchange requires higher-capacity optical communication links. Mode division multiplexing (MDM) is considered as a promising technology to support the higher data throughput. In an MDM system, the mode generator and sorter are the backbone. However, most of the current schemes lack the programmability and universality, which makes the MDM link susceptible to the mode crosstalk and environmental disturbances. In this paper, we propose an intelligent multimode optical communication link using universal mode processing (generation and sorting) chips. The mode processor consists of a programmable 4 × 4 Mach Zehnder interferometer (MZI) network and can be intelligently configured to generate or sort both quasi linearly polarized (LP) modes and orbital angular momentum (OAM) modes in any desired routing state. We experimentally establish a chip-to-chip MDM communication system. The mode basis can be freely switched between four LP modes and four OAM modes. We also demonstrate the multimode optical communication capability at a data rate of 25 Gbit/s. The proposed scheme shows significant advantages in terms of universality, intelligence, programmability and resistance to mode crosstalk, environmental disturbances, and fabrication errors, demonstrating that the MZI-based reconfigurable mode processor chip has great potential in long-distance chip-to-chip multimode optical communication systems.

We propose and experimentally demonstrate an efficient on-chip thermo-optic (TO) switch based on a photonic crystal nanobeam cavity (PCNC) and a hydrogen-doped indium oxide (IHO) microheater. The small mode volume of the PCNC and the close-range heating through the transparent conductive oxide IHO greatly enhance the coupling between the thermal field and the optical field, increasing the TO tuning efficiency. The experimental results show that the TO tuning efficiency can reach 1.326 nm/mW. And the rise time and fall time are measured to be 3.90 and 2.65 μs, respectively. In addition, compared with the conventional metal microheater, the measured extinction ratios of the switches are close (25.8 dB and 27.6 dB, respectively), indicating that the IHO microheater does not introduce obvious insertion loss. Our demonstration showcases the immense potential of this TO switch as a unit device for on-chip large-scale integrated arrays.

Abstract Optical beating is the usual approach to generation of microwave signals. However, the highest frequency achievable for microwave signals is limited by the bandwidths of optoelectronic devices. To maximize the microwave frequency with a limited bandwidth of a photodetector (PD) and relieve the bandwidth bottleneck, we propose to generate microwave signals with the single sideband (SSB) format by beating a continuous wave (CW) light with an optical SSB signal. By simply adjusting the frequency difference between the CW light and the carrier of the optical SSB signal, the frequency of the generated microwave SSB signal is changed correspondingly. In the experiment, amplitude shift keying (ASK) microwave signals with the SSB format are successfully generated with different carrier frequencies and coding bit rates, and the recovered coding information agrees well with the original pseudo random binary sequence (PRBS) of 27 − 1 bits. The proposed approach can significantly relieve the bandwidth restriction set by optoelectronic devices in high-speed microwave communication systems. Graphical Abstract

Abstract Second-order (χ (2)) optical nonlinearity is one of the most common mechanisms for modulating and generating coherent light in photonic devices. Due to strong photon confinement and long photon lifetime, integrated microresonators have emerged as an ideal platform for investigation of nonlinear optical effects. However, existing silicon-based materials lack a χ (2) response due to their centrosymmetric structures. A variety of novel material platforms possessing χ (2) nonlinearity have been developed over the past two decades. This review comprehensively summarizes the progress of second-order nonlinear optical effects in integrated microresonators. First, the basic principles of χ (2) nonlinear effects are introduced. Afterward, we highlight the commonly used χ (2) nonlinear optical materials, including their material properties and respective functional devices. We also discuss the prospects and challenges of utilizing χ (2) nonlinearity in the field of integrated microcavity photonics. Graphical Abstract

Abstract Microwave photonic filters have been regarded as an alternative to traditional radio‐frequency filters because of their wide bandwidth and large tunability. Integrated microwave photonic filters can integrate all necessary components into a single chip and are highly demanded for future radio‐frequency applications. Here, a highly integrated frequency‐tunable microwave photonic bandpass filter based on a silicon platform is proposed and demonstrated. The integrated filter consists of a phase modulator, four cascaded microring resonators and a photodetector. The frequency‐tunable range of the integrated filter is from 6.1 to 35.9 GHz, and the reconfigurable bandwidth is from 0.22 to 0.54 GHz. A large spurious free dynamic range of 102.1 dB Hz2/3 is obtained. This highly integrated approach holds great promise for miniaturised, flexible, and high‐performance microwave signal processing in modern radar and communication systems.

Density-Based Spatial Clustering of Applications with Noise (DBSCAN) is a classic density-based clustering method that can identify clusters of arbitrary shapes in noisy datasets. However, DBSCAN requires two input parameters: the neighborhood distance value (Eps) and the minimum number of sample points in its neighborhood (MinPts), to perform clustering on a dataset. The quality of clustering is highly sensitive to these two parameters. This paper introduces a parameter-adaptive DBSCAN clustering algorithm based on the Whale Optimization Algorithm (WOA-DBSCAN) to tackle this issue. The algorithm determines the parameter range based on the dataset distribution and utilizes the silhouette coefficient as the objective function. It iteratively selects the two input parameters of DBSCAN within the parameter range using the WOA. This approach ultimately achieves adaptive clustering of DBSCAN. Experimental results on five typical artificial datasets and six real UCI datasets demonstrate the effectiveness of the proposed WOA-DBSCAN algorithm. Compared with DBSCAN and its related optimization algorithms, WOA-DBSCAN shows significant improvements. The F-values of WOA-DBSCAN increased by 9.8%, 13.2%, and 2%, respectively, in two-dimensional artificial datasets. Additionally, the accuracy values on low to medium dimensional real datasets increased by 22.3%, 10%, and 23.3%. Hence, WOA-DBSCAN can maintain the clustering ability of DBSCAN while achieving adaptive parameter clustering.

Optical parametric amplifiers (OPAs) are optical amplifiers based on four-wave mixing processes, showing great potential for applications in communications, optical signal processing, quantum optics, etc. In recent years, significant progress has been made to integrated OPAs based on highly nonlinear micro-ring resonators (MRRs), benefiting from the greatly enhanced optical-matter interactions. Notable parametric gain becomes available at unprecedented low power levels, allowing for example the source-integrated coherent optical frequency combs. Therefore, an analytical formula for OPAs in high-quality (Q) MRRs is of great importance in the design of optical parametric devices. Analytical theory for OPAs in high-Q MRRs in the high pump power scenario remains elusive, where intensity-dependent nonlinear phase that brings significant parametric gains cannot be ignored. In this work, analytical formulas for the parametric gain and conversion efficiency (CE) of a high-Q MRR with high-pump excitation are derived. We show the interplay between parametric gain and field enhancement: the field enhancement of the signal and idler wave can be greatly boosted due to the compensation of the round-trip loss by the parametric gain, which in turn leads to increased field enhancement as well as greater parametric gain and CE. Our theory agrees well with numerical and experimental results.

Real-time spectral measurement technology is of great importance in observing transient phenomena and revealing ultrafast dynamics. This paper presents an approach for real-time observing intracavity soliton evolution processes in a mode-locked fiber laser by synchronizing multi-port time-division multiplexed measurements, pointing towards possible studies of intracavity dynamics in various optical systems.

Dynamic localization is a method of confining light. Here the authors demonstrate higher-order dynamic localization of photons in a synthetic temporal mesh lattice and discuss the idea of tunable temporal cloaking by combining different-order localizations.

Abstract Optical Kerr solitons generation based on microresonators is essential in nonlinear optics. Among various soliton generation processes, the single soliton generation plays a pivotal role since it ensures rigorous mode-locking on each comb line whose interval equals the free spectral range (FSR) of the microresonator. Current studies show that single soliton generation is challenging due to cavity instability. Here, we propose a new method to greatly improve single soliton generation probalility in the anomalous group velocity dispersion (GVD) regime in a micro-ring resonator based on silicon nitride. The improvement is realized by introducing mode depletion through an integrated coupled filter. It is convenient to introduce controllable single mode depletion in a micro-ring resonator by adjusting the response function of a coupled filter. We show that spectral mode depletion (SMD) can significantly boost the single soliton generation probability. The effect of SMD on the dynamics of optical Kerr solitons generation are also discussed. The proposed method offers a straightforward and simple way to facilitate robust single soliton generation, and will have an impact on the research development in optical Kerr soliton generation and on-chip optical frequency mode manipulation. Graphical abstract

Large-scale silicon-based integrated artificial neural networks lack of silicon-integrated optical neurons. Here, Yu et al, report a self-monitored all-optical neural network enabled by nonlinear germanium-silicon photodiodes, making the photonic neural network more versatile and compact.

ABSTRACT: As an indispensable part to compensate for the signal crosstalk in fiber communication systems, conventional digital multi-input multi-output (MIMO) signal processor is facing the challenges of high computational complexity, high power consumption and relatively low processing speed. The optical MIMOenables the best use of light and has been proposed to remedy this limitation. However, the currently existing optical MIMO methods are all restricted to the spatial dimension, while the temporal dimension is neglected. Here, an on-chip spatial-temporal descrambler with four channels were devised and its MIMO functions were experimentally verified simultaneously in both spatial and temporal dimensions. The spatial crosstalk of single-channel descrambler and four-channel descrambler is respectively less than -21 dB and -18 dB, and the time delay is simultaneously compensated successfully. Moreover, a more universal model extended to mode-dependent loss and gain (MDL) compensation was further developed, which is capable of being cascaded for the real optical transmission system. The first attempt at photonic spatial-temporal descrambler enriched the varieties of optical MIMO, and the proposed scheme provided a new opportunity for all-optical MIMO signal processing.

Dual-comb spectroscopy has enabled new approaches for optical precision measurements. Although Doppler-limited resolution can be achieved over long-time scales across a large bandwidth, the development of dual-comb spectroscopy is hindered by strict demands for light source stability. Typically, expensive and complex self-reference systems are required to lock the carrier-envelope offset frequency (fceo) of the laser. Additionally, simply locking the repetition frequency (frep) to a radio frequency reference source still results in residual relative timing jitter between light sources. Here we extracted the relative fceo fluctuation between the frep-locked lasers from the high-precision passive notch filtering characteristics of the phase-shifted fiber Bragg grating and then eliminated it through online phase calibration. By introducing a passive broadband Fabry–Perot cavity with excellent thermal wavelength stability, we subsequently corrected residual relative timing jitter with online wavelength calibration, and the standard deviation of the relative wavelength drift was reduced to less than 0.4 pm within the full operating range. The spectral profile can also be extracted and removed by the Fabry–Perot cavity through intensity calibration. By calibrating these three dimensions, we built a reference-free post-calibration dual-comb spectroscopy and used this powerful tool to measure the Fabry–Perot cavity resonance peaks, the notch filtering narrow band of phase-shifted fiber Bragg gratings, and the absorption characteristics of hydrogen cyanide gas. The system achieves a spectral resolution of 0.8 pm over a bandwidth of more than 100 nm. This low-cost and convenient scheme provides new ideas for the application of dual-comb spectroscopy systems.

Vessel trajectory data are currently the most important data source for vessel trajectory data mining research. However, vessel AIS data have a short sampling time interval and a large amount of data redundancy, which hampers the efficient utilization of AIS data. In order to effectively remove redundant information from AIS data and improve its usage efficiency, a compression algorithm for vessel trajectory data compression algorithm considering critical region identification (VATDC_CCRI) is proposed. The VATDC_CCRI algorithm identifies the critical regions of a vessel’s trajectory by analyzing the distribution of node variation rates. It employs the Douglas–Peucker (DP) algorithm to compress the data in these critical regions, reducing the distortion of the trajectory after compression. Additionally, the algorithm utilizes a sliding window approach to process the initial trajectory to improve the quality of the compressed vessel trajectories and retain as many spatiotemporal characteristics of the original trajectories as possible. It combines the feature nodes from the crucial regions in the vessel’s trajectory with the results obtained from the sliding window algorithm, effectively compressing the vessel’s trajectory. Experiments conducted on individual and multiple trajectories demonstrate that the VATDC_CCRI algorithm achieves higher compression rates and exhibits faster processing speeds compared to other classical vessel trajectory compression algorithms while preserving the shape of the vessel’s trajectory significantly.

The Mach–Zehnder interferometer (MZI) mesh, a mainstream structure for optical matrix-vector multiplication (MVM), has been widely employed in recently developed optical neural networks (ONNs) and combination optimization problem solvers. The conventional MZI mesh was designed specifically for complex-valued optical MVM. The network includes 2N2 phase shifters, and coherent detection is indispensable for retrieving the output complex-valued vectors. Nonetheless, the majority of applications, including ONNs, merely require real-valued optical matrices with N2 degrees of freedom (DOFs). The DOF gap between the 2 types of matrices results in a severe redundancy in the number of phase shifters when the conventional MZI mesh is applied to implement real-valued optical MVM. In this study, we propose a simplified MZI mesh for performing real-valued incoherent optical MVM. It has N2 phase shifters and an optical depth of N + 1, and it outperforms the conventional MZI mesh. Furthermore, we constructed an ONN with the proposed MZI mesh and successfully performed the iris classification task via in situ training of particle swarm optimization. More importantly, we introduced a matched on-chip nonlinear activation function, so the proposed MZI mesh can be cascaded onto a single chip. Overall, the proposed real-valued MZI mesh and in situ training method are space efficient, energy efficient, scalable, and robust to fabrication errors. Therefore, they are suitable for large-scale ONNs.

This review summarizes the advances of photonic accelerators from the viewpoint of photonic matrix multiplication, providing a guidance for all-optical or optoelectronic-hybrid AI hardware chip system.

Abstract Background To retrospectively compare clinical and radiological results of long-segment fixation (LF) and six-screw short-segment fixation combined with kyphoplasty (SSFK) for osteoporotic thoracolumbar burst fracture (OTBF). Methods Forty patients affected by OTBF with mean age of 61.85 years were included in this study. The mean follow-up period was 13.63 months. Twenty-four patients were treated by SSFK, and 16 patients were treated by LF. Clinical outcomes, radiological parameters and complications were assessed and compared. Results The mean operative time and blood loss were 89.71 ± 7.62 min and 143.75 ± 42.51 ml for SSFK group, respectively; 111.69 ± 12.25 min (P

Abstract Background To observe the clinical efficacy of an anterior single rob-screw fixation (ASRSF) combined with the oblique lumbar intervertebral fusion (OLIF) approach compared with a posterior percutaneous screw fixation (PPSF) combined with OLIF in the treatment of lumbar spondylolisthesis. Method This is a retrospective case–control study. Patients with degenerative lumbar spondylolisthesis (DLS) treated with either ASRSF combined with OLIF or PPSF combined with OLIF from January 2016 to January 2018 were enrolled in this study. None of the patients had posterior decompression. The visual analog scale (VAS) and Oswestry dysfunction index (ODI) were used for clinical efficacy assessment. The pre- and post-operational disc height, height of foramen, subsidence, and migration of cages, fusion rate and surgery-related complications were compared between the two groups. Results Fifty-three patients were included in this single-center study. According to the fixation methods, patients were divided into the ASRSF group (group A, 25 cases) and the PPSF group (group B, 28 cases). There was no statistical difference in surgery-related complications between groups. There was a significant difference in the VAS score at one-week post-surgery (2.3 ± 0.5 vs. 3.5 ± 0.4, P = 0.01), and three months post-operation (2.2 ± 0.3 vs. 3.0 ± 0.3, P = 0.01). Comparison of post-operative imaging data showed that there was a significant difference in the height of the foramen between groups at three months post-surgery(18.1 ± 2.3 mm vs. 16.9 ± 1.9 mm, P = 0.04). At 24 months post-surgery, the ODI was 12.65 ± 3.6 in group A and 19.1 ± 3.4 in group B (P = 0.01). Twelve months after surgery, the fusion rate in group A at 72.0% and 78.6% in group B was not statistically significant (P = 0.75). Fusion was identified in all patients at 24 months post-surgery. Conclusion When compared to PPSF, ASRSF combined with OLIF for DLS can reduce post-operative low back pain in the initial stages, maintain the height of the foramen and improve the performance of lumbar function.

All-optical control of silicon photonic devices plays a key role in on-chip all-optical computing, switching and signal processing. However, due to the weak intrinsic nonlinear effects of silicon, current integrated devices are limited by high power consumptions. Here, an all-optical controllable optomechanical microring resonator (OMMR) is proposed and fabricated on the silicon platform. Due to the mechanical Kerr effect induced by the optical gradient force (OGF), we realize seamless wavelength tuning over a range of 2.56 nm which is 61% of the free-spectral range, with a high tuning efficiency of 80 GHz/mW. By controlling the pump power, the OMMR indicates three working regions, which are the cutoff region, amplified region and saturate region. Accordingly, this function is similar to a transistor. Furthermore, the dynamic properties of the OMMR are investigated. By analyzing the dynamic responses, we demonstrate that the OMMR is driven by the OGF, rather than other nonlinear effects. The method simplifies the experimental setup, compared with measuring the intrinsic mechanical frequency of the OMMR. This all-silicon device is energy-efficient and compatible with complementary metal-oxide semiconductor (CMOS) processing. We believe that this work is important for on-chip all-optial signal processing and beneficial to futher studies on integrated optomechanics.

Abstract In recent years, with the in-depth research on spinal tuberculosis, posterior surgery alone has been praised highly by more and more surgeons due to the better correction of kyphosis, better maintenance of spinal physiological curvature, smaller surgical trauma and fewer surgical complications. However, there is currently lack of relevant reports about the efficacy of posterior surgery alone in the treatment of tuberculosis in the T4–6 segments. This study aimed to evaluate the clinical study efficacy and feasibility of one-stage posterior-only surgical treatment for thoracic spinal tuberculosis in the T4–6 segments. 67 patients with tuberculosis in T4–6 segments who underwent one-stage posterior-only surgery were included in this study. The clinical efficacy was evaluated using statistical analysis based on the data about erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), Oswestry Dability Index (ODI) score, Visual Analogue Scale (VAS) score and Cobb angle before surgery, after surgery and at the last follow-up. All patients completed fusion during the follow-up period of 6–9 months. ESR and CRP were returned to normal for all patients at 6 months follow-up. In the meanwhile, among the 27 patients combined with neurological impairment, neurological functions of 22 cases (81.48%) recovered completely at the last follow-up (P

Microwave photonics is a promising and rapidly developing interdisciplinary field. It combines microwave and photonic techniques to generate, transmit, process, and manipulate microwave signals by using the advantages of broadband, high frequency, and low loss provided by photonics. As an important branch of microwave photonics, the microwave photonic filter (MPF) can overcome the limitations set by traditional electronic technology and can realize advanced signal processing in modern communication systems due to its higher performance, selectivity, and flexibility. This review provides a comprehensive overview of MPFs, including fundamental principles, typical structures, and key applications. Additionally, the microwave photonic integration is a very important tendence because of its advantages of small size, light weight, low power consumption, and low cost. The recent advances in integrated MPF are also reviewed.

Along with the known risk factors of cardiovascular diseases (CVDs) constituting metabolic syndrome (MS), the gut microbiome and some of its metabolites, in particular trimethylamine-N-oxide (TMAO), are actively discussed. A prolonged stay under natural hypoxic conditions significantly and multi-directionally changes the ratio of gut microbiome strains and their metabolites in feces and blood, which is the basis for using hypoxia preconditioning for targeted effects on potential risk factors of CVD. A prospective randomized study included 65 patients (32 females) with MS and optimal medical therapy. Thirty-three patients underwent a course of 15 intermittent hypoxic–hyperoxic exposures (IHHE group). The other 32 patients underwent sham procedures (placebo group). Before and after the IHHE course, patients underwent liver elastometry, biochemical blood tests, and blood and fecal sampling for TMAO analysis (tandem mass spectrometry). No significant dynamics of TMAO were detected in both the IHHE and sham groups. In the subgroup of IHHE patients with baseline TMAO values above the reference (TMAO ≥ 5 μmol/l), there was a significant reduction in TMAO plasma levels. But the degree of reduction in total cholesterol (TCh), low-density lipoprotein (LDL), and regression of liver steatosis index was more pronounced in patients with initially normal TMAO values. Despite significant interindividual variations, in the subgroup of IHHE patients with MS and high baseline TMAO values, there were more significant reductions in cardiometabolic and hepatic indicators of MS than in controls. More research is needed to objectify the prognostic role of TMAO and the possibilities of its correction using hypoxia adaptation techniques.

The great success of artificial intelligence (AI) calls for higher-performance computing accelerators, and optical neural networks (ONNs) with the advantages of high speed and low power consumption have become competitive candidates. However, most of the reported ONN architectures have demonstrated simple MNIST handwritten digit classification tasks due to relatively low precision. A microring resonator (MRR) weight bank can achieve a high-precision weight matrix and can increase computing density with the assistance of wavelength division multiplexing (WDM) technology offered by dissipative Kerr soliton (DKS) microcomb sources. Here, we implement a car plate recognition task based on an optical convolutional neural network (CNN). An integrated DKS microcomb was used to drive an MRR weight-bank-based photonic processor, and the computing precision of one optical convolution operation could reach 7 bits. The first convolutional layer was realized in the optical domain, and the remaining layers were performed in the electrical domain. Totally, the optoelectronic computing system (OCS) could achieve a comparable performance with a 64-bit digital computer for character classification. The error distribution obtained from the experiment was used to emulate the optical convolution operation of other layers. The probabilities of the softmax layer were slightly degraded, and the robustness of the CNN was reduced, but the recognition results were still acceptable. This work explores an MRR weight-bank-based OCS driven by a soliton microcomb to realize a real-life neural network task for the first time and provides a promising computational acceleration scheme for complex AI tasks.

Enterogenous cyst (EC) is a rare congenital lesion generally located in the central nervous system, such as in the cerebral hemispheres, posterior fossa, or spinal canal. They are usually benign lesions, and malignant transformation is rare. A 42-year-old woman felt an obvious pain in the lump and went to a local hospital for local lumpectomy. After 7 months, she again felt pain in the buttocks and difficulty in urinating and defecation. The computed tomography (CT) scan showed a mass in the pelvis. Sacrococcygeal cyst excision was performed 10 days later, and postoperative pathology showed epidermoid cyst. Shortly after, the patient recovered and was discharged from the hospital; the pain in the buttocks continued to recur. Puncture and drainage were performed five times. Later, the patient went to our hospital for treatment, and pelvic MRI showed multiple abnormal signal shadows in the presacral and sacrococcygeal regions, some of which were considered abscesses, and some were cystic lesions. She underwent tumor resection and was diagnosed with EC with locally moderately differentiated adenocarcinoma. Four months later, the patient’s symptoms of swelling and pain recurred. MRI examination showed multiple high-signal T2 shadows in the anterior sacral and subcutaneous tissues of the buttocks, and enhanced scan showed partial marginal enhancement. After assessment, the patient was given a radiation dose of 60 Gy/25F. ECs in the anterior sacral and soft tissue of the buttocks are very rare, and the case of carcinomatous transformation has never been reported. Therefore, we discussed the clinicopathological features of ectopic ECs and reviewed the literature.

The capability of characterizing arbitrary and non-repetitive emission spectra with a high resolution in real-time is of great merit in various research fields. Optical frequency combs provide precise and stable frequency grids for the measurement of a single spectral line with high accuracy. Particularly, dual-comb spectroscopy enables spectral measurement with a large bandwidth spanning tens of nanometers, but it is limited to measuring absorption spectra and has to trade-off spectral resolution vs the acquisition frame rate set by the repetition rate. Here, to alleviate these restrictions, we propose and demonstrate time-resolved spectroscopy for an emission spectrum based on hybrid asynchronous optical sampling, which features a spectral resolution of 0.63 pm, a frame rate of 1 MHz, and a measurement bandwidth of 13.6 nm, simultaneously. A mode-locked fiber comb with a repetition frequency of f1 is harnessed to interrogate emission spectral features with high resolution via optical Fourier transform achieved using a time-lens. Subsequently, a soliton microcomb of repetition frequency f2s ≈ 1000f1 serving as a probe pulse implements hybrid asynchronous optical sampling, thus significantly increasing the acquisition rate by nearly 3 orders of magnitude. As a proof-of-concept demonstration, the frequency trajectory of a rapidly scanning laser with a linear chirp of 6.2 THz/s is tracked. We believe that chip-scale microcombs will make the fast and high-resolution emission spectroscopy presented here a powerful tool for widespread applications.

Nonreciprocal devices are crucial in scientific research and practical applications at all frequencies. Here the authors demonstrate an integrated terahertz optical isolator based on the magneto-optical effect in a nonreciprocal resonator.

Graphene-based photodetectors have many advantages for applications. Here, the authors demonstrate a high-speed optical coherent receiver for optical communications based on graphene-on-plasmonic slot waveguide photodetectors.

Material loss, especially in metal and 2D materials, is the bottleneck for high‐performance integrated optical devices. Here, a novel concept by utilizing the high‐loss materials in a non‐Hermitian system instead of avoiding the loss materials. It is theoretically analyzed how the complex refractive index affects the loss of optical devices based on a two‐waveguide‐coupled non‐Hermitian system. The results reveal that the loss of optical devices can be improved by increasing the loss of materials or the difference in real refractive indexes. It is experimentally verified that a high loss material can be placed close to the optical waveguide without introducing non‐negligible loss to the optical devices, by demonstrating a thermo‐optic phase shifter with a metallic heater placed very close to the silicon waveguide. As a result, the insertion loss is only 0.1 dB for the 100 μm long heater with a gap of 400 nm, and the largest bandwidth for the metallic heater reaches up to 280 kHz. The finding offers a way to realize high‐performance optical devices with high‐loss materials, contributing to the practical applications of strongly absorbing materials based on non‐Hermitian optics.

The cascaded deep-learning network of YOLOv3 emphasizes on the layer-wise feature extraction. It neglects the sequential influence among the layers that contributes to the subtle features for the objects detection. An improved YOLOv3 model with skipping connections is proposed in this paper for the sufficient utilization of layer-wise features. Firstly, a DenseBlock network is adopted as the fourth and fifth down-sampling layers of YOLOv3. The DenseBlock is characterized of a parallel architecture and capable of transmitting the features backwards among extraction layers. Then, the features of preceding layers are incorporated by a skipping fashion into subsequential layers. Secondly, a spatial pyramid pooling (SPP) module is introduced at the neck of the object detector. It realizes the size-tuning of the model input. Then the multi-scaled region features are generated after the pooling and concatenation operation of the SPP. Finally, the validation experiments have been conducted on a dataset of helmet objects. The results have shown that the proposed YOLOv3 model improves the accuracy effectively. It yields a mean average precision 88.6% on the helmet detection, which is 3.5% higher than the original YOLOv3 network.

The fault detection and diagnosis of a gas turbine is of great significance for guaranteeing the complicated dynamic systems working normally and safely. Most of the existing fault diagnosis methods, based on convolutional neural networks (CNN), have certain limitations in extracting correlations of multi-channel data features. The accuracy of fault diagnosis still needs to be improved. In this paper, an approach of fault diagnosis, based on matrix capsules with EM routing, is presented. First of all, three channels data, which respectively represent acceleration, pressure and pulse, are integrated into one image to feed into the network. Secondly, network models based on the matrix capsules start to be trained by using input dataset which contains fault image and normal image. Finally, the pre-trained capsules model is used to diagnose the state of testing data. Besides, to verify the superiority of the algorithm used in this paper, a comparative experiment is implemented between matrix capsule networks and CNN. The results demonstrate that the testing accuracy is 99.995%.

Voxel grid is widely used in point cloud segmentation due to its regularity. However, the memory consumption caused by high resolution restricts the performance of voxel grid. This paper proposes an improved voxel grid deep network (IVDN) model to represent more comprehensive point cloud features at the same resolution, thus improving the segmentation performance of point cloud. Firstly, the point cloud data are structured within a voxel bounding box to correspond with the three-dimensional(3D) convolution kernel, and a fixed number of point coordinates are selected to generate the point feature vector. Then, in order to consider the distribution characteristics, the reliability coefficient is used as an equivalent descriptor of the point cloud distribution density. Finally, a corresponding deep network is constructed to deal with the above features. Experimental results show that the proposed IVDN model can improve the mean classification accuracy and segmentation index mIoU(Mean Intersection over Union) effectively, with a 0.45% and 0.3% improvement on Shape net16 dataset respectively.

Obesity and associated chronic inflammation were shown to facilitate breast cancer (BC) growth and metastasis. Leptin, adiponectin, estrogen, and several pro-inflammatory cytokines are involved in the development of obesity-driven BC through the activation of multiple oncogenic and pro-inflammatory pathways. The aim of this study was to assess the reported mechanisms of obesity-induced breast carcinogenesis and effectiveness of conventional and complementary BC therapies. We screened published original articles, reviews, and meta-analyses that addressed the involvement of obesity-related signaling mechanisms in BC development, BC treatment/prevention approaches, and posttreatment complications. PubMed, Medline, eMedicine, National Library of Medicine (NLM), and ReleMed databases were used to retrieve relevant studies using a set of keywords, including “obesity,” “oncogenic signaling pathways,” “inflammation,” “surgery,” “radiotherapy,” “conventional therapies,” and “diet.” Multiple studies indicated that effective BC treatment requires the involvement of diet- and exercise-based approaches in obese postmenopausal women. Furthermore, active lifestyle and diet-related interventions improved the patients’ overall quality of life and minimized adverse side effects after traditional BC treatment, including postsurgical lymphedema, post-chemo nausea, vomiting, and fatigue. Further investigation of beneficial effects of diet and physical activity may help improve obesity-linked cancer therapies.

Dissipative Kerr solitons in microcavity systems exhibit remarkable nonlinear dynamics. The real-time measurement of soliton motion facilitates the comprehensive understanding of soliton physics. In this Letter, an all-optical radio frequency (RF) spectrum analyzer (named f-LISA) is used to characterize various stable soliton states and to track relative soliton motion in real time. By applying an inverse Fourier transform to the broadband RF spectrum, the autocorrelation traces are obtained with a temporal resolution of 373 fs and an ultrahigh frame rate of 20.6 MHz. We successfully characterize not only the stable single soliton state but also the stable multi-soliton states with different azimuthal angles between adjacent solitons. Furthermore, the dynamics of soliton switching from four-soliton state to single soliton state is observed in a temporal window of 60 µs. It is believed that the proposed scheme provides an alternative way to visualize the multi-soliton trajectories and enable the study of the soliton dynamics in integrated microcavities.

Abstract Background Pre-existing degeneration of adjacent segment is an important risk factor for adjacent-segment degeneration (ASD), but only limited and controversial studies have addressed its management. Methods We retrospectively analyzed patients with symptomatic degeneration of the L5/S1 segment warranting surgical interference and severe asymptomatic degeneration of the L4/5 segment. Of these patients, those who underwent interbody fusion of the causative (L5/S1) segment and distraction of the intervertebral space and facet fusion of the adjacent L4/5 segment were included in Group A (n = 103), while those who underwent interbody fusion of both the L5/S1 and L4/5 segments were included in Group B (n = 81). Clinical and radiographic outcomes were evaluated. Results Mean follow-up time was 58.5 months (range, 48–75 months). We found no significant difference in clinical outcomes or incidence of ASD in the L3/4 segment between Groups A and B. Compared with Group B, Group A experienced less bleeding (315 ± 84 ml vs. 532 ± 105 ml), shorter operation time (107 ± 34 min vs. 158 ± 55 min) and lower costs (US $13,830 ± $2640 vs. US $16,020 ± $3380; P

Abstract Bloch oscillations (BOs) were initially predicted for electrons in a solid lattice to which a static electric field is applied. The observation of BOs in solids remains challenging due to the collision scattering and barrier tunnelling of electrons. Nevertheless, analogies of electron BOs for photons, acoustic phonons and cold atoms have been experimentally demonstrated in various lattice systems. Recently, BOs in the frequency dimension have been proposed and studied by using an optical micro-resonator, which provides a unique approach to controlling the light frequency. However, the finite resonator lifetime and intrinsic loss hinder the effect from being observed practically. Here, we experimentally demonstrate BOs in a synthetic frequency lattice by employing a fibre-loop circuit with detuned phase modulation. We show that a detuning between the modulation period and the fibre-loop roundtrip time acts as an effective vector potential and hence a constant effective force that can yield BOs in the modulation-induced frequency lattices. With a dispersive Fourier transformation, the pulse spectrum can be mapped into the time dimension, and its transient evolution can be precisely measured. This study offers a promising approach to realising BOs in synthetic dimensions and may find applications in frequency manipulations in optical fibre communication systems.

Non-orthogonal multiple access (NOMA) is widely recognized to increase the number of users and enhance the spectral efficiency in fifth-generation (5G) wireless networks and beyond. NOMA is still in the theoretical analysis and simulation phases and fewer experimental works are reported to date. In this paper, we design and implement NOMA in software-defined radio, and evaluate its performance. This includes the real-time realization of the key components of NOMA, i.e., superposition and successive interference cancellation. The main novelty of this paper is to introduce constructing superimposed signals with varying symbol rates to enlarge the achievable rate region of the experimented NOMA. By applying varying symbol rates, the set of possible transmission rate pairs enlarges and we can reach higher data rates compared to existing modulation and coding schemes (MCS). We also propose an algorithm to efficiently find the rate pairs. Simulations and experiments demonstrate that NOMA with a varying symbol rate not only can reach higher data rates than orthogonal schemes such as time division multiple access, but it can also outperform existing MCS-based methods which have a fixed symbol rate. The experiments also show that there is a noticeable gap between NOMA in theory and practice. In addition to the new NOMA experiments, we review the state-of-the-art in experimental NOMA. We also discuss several directions for future experiments that can help bridge the gap between theory and practice and bring NOMA to practical communication systems.

Graphene silicon devices have attracted significant attention due to their outstanding electronic and optical properties. By electrically tuning the Fermi level of the monolayer graphene sheet, electro-optic (EO) modulators have been widely analyzed. Despite significant progress has been achieved on high-speed modulation with integrated graphene silicon waveguides, it remains challenging to realize a high modulation efficiency since the light absorption in graphene is limited. Here, we propose and experimentally demonstrate a high efficiency EO modulator using a graphene-assisted tapered silicon ring resonator, where the tapered zone is covered by a graphene/graphene capacitor. This graphene-assisted tapered zone strongly enhances the graphene-light interaction that the achieved static and dynamic modulation (1KHz) depths are up to 26 dB and 12 dB, respectively. Physical mechanism of the modulation based on this nanostructure is also discussed in detail. The proposed nanostructure provides a promising alternative method for high-performance EO modulator, which is essential for the on-chip optical communication, optical computing and optical signal processing.

Photonic integrated circuits (PICs) have been a research hotspot in recent years. Programmable PICs that have the advantages of versatility and reconfigurability that can realize multiple functions through a common structure have been especially popular. Leveraging on-chip couplers and phase shifters, general-purpose waveguide meshes connected in different topologies can be manipulated at run-time and support a variety of applications. However, current waveguide meshes suffer from relatively a low cell amount and limited bandwidth. Here, we demonstrate a reconfigurable photonic integrated computing chip based on a quadrilateral topology network, where typical analog computing functions, including temporal differentiation, integration, and Hilbert transformation, are implemented with a processing bandwidth of up to 40 GHz. By configuring an optical path and changing the splitting ratio of the optical switches in the network, the functions can be switched and the operation order can be tuned. This approach enables wideband analog computing of large-scale PICs in a cost-effective, ultra-compact architecture.

Abstract Polaritons in two‐dimensional (2D) materials continues to garner significant attention due to their favorable ability of field‐confinement and intriguing potential for low‐loss and ultrafast optical and photonic devices. The recent experimental observation of in‐plane anisotropic dispersion in natural van der Waals materials has revealed much richer physics as compared to isotropic plasmonic materials, which provides new insight to manipulate the polaritons and manufacture flat optical devices with unprecedented controls. Herein, we give an overview of the recent progress in in‐plane anisotropic polaritons launched and visualized in the near‐field range in 2D layered van der Waals materials. Furthermore, future prospects in this promising but emerging field are featured on the basis of its peculiar applications. This review article will stimulate the scientific community to explore other hyperbolic materials and structures in order to develop optical technologies with novel functionalities and further improve the understanding of the exotic photonic phenomena.

In the absence of random mode mixing (RMM), linearization of the cross-mode modulation (XMM) under pump-probe configuration is in general complicated because the evolution of the pump light depends on the local mode decomposition of itself. In this work, general derivation of the Hamiltonian of the XMM system without RMM is carried out and discussed under two interesting scenarios where the pump evolution can be effectively linearized, leading to spatial dependent Hamiltonian of the probe light. Representative evolutions of pump and probe light over transmission are investigated with the assistance of Poincaré sphere based on the Hamiltonian approach. Our results are benchmarked against the results given by precession equations examined by Lin and Agrawal and numerical simulations using nonlinear coupled mode equations (CMEs). By highlighting the eigenstates as well as eigenvalues of the system, our approach provides an intuitive yet powerful approach to understand the XMM nonlinear problems and to dynamically manipulate the spatial profiles of the probe light.

A generalized Lugiato-Lefever equation in frequency domain for the modeling of Kerr frequency combs, including full-order frequency dependence of all parameters, is derived and solved without sacrifice of computational time. The proposed model shows high accuracy and efficiency in simulations, agreeing well with previous reports. Based on the proposed model, the influence of parametric frequency dependence is systemically investigated by generating octave-spanning combs in a silicon-nitride micro-ring-resonator with an air slot. The device is dispersion-tailored for broadband comb generation, and its parameters are strongly dependent on frequency. When considering three kinds of parametric influence separately, we reach the following conclusions. The frequency dependent propagation constant determines whether the dispersive wave is generated, matching well with the phase-matching condition. Besides, the frequency dependence of effective area and nonlinear refractive index can both introduce spectral blue-shift. Combining full-order frequency dependence of the three parameters together, a dispersive wave sideband emerges and this phenomenon has never been reported. Our work reveals that full-order frequency dependence of all parameters can be considered with high computational efficiency, which enables precise prediction of dispersive wave frequency for f-2f self-referencing and opens up more reliable avenues for the understanding of Kerr frequency comb generation, especially octave-spanning cases.

Highly nonlinear multimode optical integrated devices hold great promise for the emerging large-capacity mode division multiplexing (MDM) technology, and they correspondingly enable multimode optical signal processing at nodes of the MDM network. Here, we propose a highly nonlinear multimode organic-silicon slot waveguide (HN-OSSW) whose nonlinearity is greatly improved by the intensive optical field confinement in the nano-gaps filled with highly nonlinear organic material. Specifically, the achieved nonlinear coefficients of the HN-OSSW under TE0 and TE1 modes are estimated higher than 7000 W-1m-1 and 5800 W-1m-1, respectively. Also, by phase mismatch optimization, this nanostructure exhibits superior intramode four-wave mixing conversion efficiency but ignorable intermode crosstalk. Coupling efficiency of the mode multiplexer and the optimal nonlinear interaction length are discussed in detail. Based on the HN-OSSW, we then numerically demonstrate a multimode all-optical hexadecimal addition and subtraction for sixteen phase-shift keying (16PSK) signals with the operation rate of 1.28Tb/s. Six different logic operations including A+B-C, A+C-B, B+C-A, A+B, A-B and B-A are simultaneously obtained with good performance evaluations, indicating the promising prospect of the HN-OSSW applied in ultrafast multimode signal processing.

Currently, high dynamic range (HDR) videos with high resolution (HR) have become popular due to the display and the rendered technological advancements. However, making ultra-high definition (UHD) with HDR videos is expensive. The legacy low-resolution (LR) standard dynamic range (SDR) format is still largely used in practice. It is necessary to search for a solution to transform LR SDR videos into UHD HDR format. In this paper, we consider joint super resolution and learning inverse tone mapping an issue of high-frequency reconstruction and local contrast enhancement, and we propose an architecture based on a generative adversarial network to apply joint SR-ITM learning. Specifically, we include the residual ResNeXt block (RRXB) as a basic module to better capture high-frequency textures and adopt YUV interpolation to achieve local contrast enhancement. By adopting a generative adversarial network as a pivotal training mechanism, our designs show advantages in both integration and performance. Our code is now available on GitHub: SR-ITM-GAN.

Holothurian counting is a basic requirement in aquaculture. An in situ noncontact approach for holothurian counting is basically harmless, which makes such techniques an optimal choice for aquaculture monitoring. To develop an in situ noncontact counting system, there are two main challenges. In the underwater environment, the light is absorbed by the water, which distorts the underwater image color. Another challenge comes from the dynamic background, which makes it difficult to associate the ID to each holothurian. In this paper, we propose an effective usual framework for holothurian counting. The framework is composed of three modules: an underwater enhancement module, a multiple object detection module and a multiple object tracking module. To evaluate our system comprehensively, we collect and release the first underwater multiple object tracking dataset following the same format as MOT-16. The results show that our system can practically detect, track and count holothurians under different conditions.

Water molecules play a very important role in the hydration and dehydration process of hydrates, which may lead to distinct physical and chemical properties, affecting their availability in practical applications. However, miniaturized, integrated sensors capable of the rapid, sensitive sensing of water molecules in the hydrate are still lacking, limiting their proliferation. Here, we realize the high-sensitivity sensing of water molecules in copper sulfate pentahydrate (CuSO4·5H2O), based on an on-chip terahertz whispering gallery mode resonator (THz-WGMR) fabricated on silicon material via CMOS-compatible technologies. An integrated THz-WGMR with a high-Q factor of 3305 and a resonance frequency of 410.497 GHz was proposed and fabricated. Then, the sensor was employed to distinguish the CuSO4·xH2O (x = 5, 3, 1). The static characterization from the CuSO4·5H2O to the copper sulfate trihydrate (CuSO4·3H2O) experienced blueshifts of 0.55 GHz/μmol, whereas the dehydration process of CuSO4·3H2O to copper sulfate monohydrate (CuSO4·H2O) exhibited blueshifts of 0.21 GHz/μmol. Finally, the dynamic dehydration processes of CuSO4·5H2O to CuSO4·3H2O at different temperatures were monitored. We believe that our proposed THz-WGMR sensors with highly sensitive substance identification capabilities can provide a versatile and integrated platform for studying the transformation between substances, contributing to hydrated/crystal water-assisted biochemical applications.

Summary: Ten lanthanacarborane complexes were synthesized to study the rare B−Hδ−∙∙∙Mn+ inverse hydrogen bonds (IHBs). The average bonding energy of B−Hδ−∙∙∙Ln3+ is theoretically determined to be larger than 24 kJ/mol, which is comparable to moderately strong hydrogen bonds (21–56 kJ/mol). In addition to NMR and IR, magnetometer was used to study the exchange-coupling interaction via such B−Hδ−∙∙∙Ln3+ IHBs in detail, and the coupling constant is determined to be −2.0 cm−1, which is strong enough to compare with single-atom bridged dysprosium(III) complexes. Two imidazolin-iminato incorporated complexes have shown energy barrier for magnetization reversal larger than 1000 K, and the exchange-biasing effects are evident. Moreover, the bonding strengths of B−Hδ−∙∙∙Mn+ IHBs are cation-dependent. If M = Na, the B−Hδ−∙∙∙Na+ bonding energy is reduced to 14 kJ/mol, and the dimerization process is no longer reversible. The exchange-biasing effect is also disappeared. We believe such a finding extends our knowledge of IHBs.

The characterization of the state of polarization is of great importance in broad applications, such as microscopy, communications, astronomy, and remote sensing. In this Letter, we propose and demonstrate a novel integrated division-of-time polarimeter (DOTP) based on a Mach–Zehnder interferometer and two photodetectors (PDs). The proposed DOTP achieves improved measuring efficiency and accuracy by measuring a pair of orthogonal polarization states simultaneously. The analysis matrix, which is used to recover the Stokes vector, is elaborately optimized to reduce the influence of the PD noise. Compared to the conventionally designed DOTP, the measuring efficiency is improved by 33% and the equally weighted variance, a figure of merit used to characterize the total variance of the Stokes vector, is also reduced by 33%. The performance of the proposed device is experimentally characterized by comparing with a commercial product. Furthermore, a method based on the least-squares method and singular value decomposition is adopted to quantize the deviation between the ideal optimal analysis matrix and the practical one.