Shaoyang Men

Introduction: Tongue diagnosis is a vital component of Traditional Chinese Medicine (TCM), with cracked tongue serving as a key diagnostic indicator. Although deep learning techniques have advanced TCM tongue image recognition, existing methods for cracked tongue detection still face challenges of low accuracy and slow processing speed. Methods: To address these issues, we propose an improved YOLOv5 framework integrating Switchable Atrous Convolution (SAC) and a Transformer encoder. First, given the varied morphology, size, and distribution of tongue cracks, SAC is introduced into the backbone network to adaptively adjust receptive fields via different atrous rates, enabling multi-scale feature extraction and better preservation of fine details. Second, a Transformer encoder module is incorporated to capture global information and long-range dependencies, thereby enhancing the network's overall feature representation capability. Experiments are conducted on a self-built tongue image dataset to evaluate the proposed modifications. Results: The improved YOLOv5 demonstrates superior performance over the original YOLOv5 in tongue crack detection, achieving an increase of 2.2% in AP@50 and 8.43% in AP@50:95 metrics on the self-built dataset. Discussion: These results confirm that the combination of SAC and Transformer effectively boosts detection accuracy and generalizability, addressing the limitations of prior approaches. The proposed method offers a more reliable and efficient solution for automated tongue crack analysis, with potential implications for clinical TCM diagnostic support.

Background: Digital transformation in Traditional Chinese Medicine (TCM) is hindered by data silos, poor standardization, and inefficient quality control. This study developed and validated an intelligent data governance system to address these gaps. Methods: Using Chest Bi syndrome (coronary heart disease) as a model, we established a standardized data system (9 subsets, 352 elements; 42.6% TCM-specific) based on the “Triple-Loop Coupling” theory and HL7 FHIR (Fast Healthcare Interoperability Resources)/SNOMED CT standards. The system integrates a three-tier trigger engine (128 rules) for real-time alerts and a domain-adapted LLM, fine-tuned on TCM texts and guidelines, for unstructured data processing. Efficacy was evaluated via a prospective controlled study ( n = 972). Conclusion: The intervention group showed significant improvements ( P < 0.001): data completeness rose from 82.51% to 99.59% ( Δ =17.08%), and TCM-specific data elements collection improved from 67.08% to 98.97% ( Δ =31.89%). Logical consistency errors dropped from 15.84% to 1.65% ( Δ =-14.20%). Unstructured data conversion increased from 45.06% to 91.77% ( Δ =46.71%). Quality control time per record decreased by 58.95% (from 12.48 to 5.12 min). The AI model achieved a 91.4% F1-score and 91.8% alert accuracy. Conclusion: The “standardization-trigger-AI” model significantly enhances data quality and efficiency for TCM. This system provides a preliminary and verifiable exploratory path for the intelligent governance of TCM clinical data, offering a scalable solution for TCM modernization.

In this work, high-resolution sparse channel estimation in multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) systems is addressed. Firstly, a block-structured compressed channel sensing (CCS) model with high spectral efficiency and high delay resolution is constructed. Then, by fully exploiting the temporal correlation and joint sparsity of the channels, a novel two-stage prior delay support-aided delay tracking and block residual norm minimization (PDSA-DT-BRNM) algorithm is proposed. In the first stage, with a limited number of pilots for each antenna and the delay grids within the prior delay support, an efficient delay tracking and block norm minimization algorithm is put forward to choose the common delay grids and estimate each block gain iteratively. In the second stage, by comprehensively utilizing the intermediate channel estimation results of the first stage and the prior delay support, an optimized channel estimation strategy is developed based on the block residual norm minimization (BRNM) criterion. Simulation results and theoretical analysis show the effectiveness of the proposed channel estimation scheme in terms of channel estimation performance, spectral efficiency and computational complexity.

The characteristics of the tongue surface and sublingual vein patterns provide valuable insights into an individual’s health status and have long served as the cornerstone of traditional tongue diagnosis. As a non-invasive digital biomarker, tongue imaging has recently gained attention as a promising modality for capturing internal physiological and pathological variations, with the potential to support remote healthcare delivery and continuous health monitoring. Nevertheless, conventional practice remains highly dependent on subjective clinical judgment, which often introduces variability in diagnostic accuracy and therapeutic decision-making. To mitigate these limitations, computerized tongue image analysis (CTIA) has been developed to enhance objectivity, reproducibility, and consistency. This review proposes a structured taxonomy of CTIA, encompassing the essential stages of image acquisition, preprocessing, dataset construction, feature extraction, and disease detection. By systematically synthesizing advances across these stages, we delineate key challenges and outline potential solutions, particularly regarding data standardization and feature quantification. The taxonomy is intended to provide a coherent framework that may contribute to improving diagnostic precision and reliability, thereby informing the gradual clinical integration of tongue imaging as a supportive tool for non-invasive disease screening.

Infant facial expression recognition is of great importance for early care, which can help parents and caregivers to identify the emotional state of infants. In this study, we propose an improved YOLOv8 model for infant facial expression recognition. Compared with traditional YOLOv8 model, following improvements have been proposed. First, the Swin Transformer V2 structure is introduced in the backbone network part to improve the model’s ability to model global features. Second, the AKConv and RepViTBlock modules are added to the head structure to optimize the feature extraction and fusion process. Then, the BiFPN is treated as the feature pyramid network (FPN) module to enhance the multi-scale feature fusion capability. Based on the above improvements, the proposed YOLOv8 model can effectively handle the subtle differences and multi-scale changes in infant facial expression, resulting in high-Precision and high-Recall facial expression recognition, which is suitable to recognize the complex and fine-grained facial expression. The applied dataset is composed of more than 9000 images, which are labeled as cry, happy, neutral and back of head. The experimental results show that the improved YOLOv8 model outperforms the original YOLOv8 and other mainstream facial expression recognition models.

Container terminal yards are automated terminal yards. The rail-mounted gantry crane (RMGC) and rail-mounted gantry crane tracks (RMGCTs) that frequently operate in terminal yards need to be inspected regularly to ensure the safe operation of container transportation in the yard. This paper proposes a framework for the path planning of RMGCT visual inspection using a UAV in container terminal yards in a port environment. The framework consisted of two main aspects. First, a global path to all inspection viewpoints was planned according to the inspection requirements using the A* algorithm and the improved minimum snap method. Second, the intelligent bidirectional rapidly exploring random trees star (IB-RRT*) algorithm was introduced to plan the local path during the flight of the UAV. Finally, the feasibility of the path-planning framework was demonstrated using simulation experiments.

Body constitution of traditional Chinese medicine(TCM) can be used to guide the prevention, diagnosis, treatment, rehabilitation and health preservation of diseases. At present, the identification of damp-heat constitution and balanced constitution is mostly determined by questionnaire, which leads to the great influence of subjective factors. Therefore, we propose an effective classification model DenseNet-CBAM for identifying damp-heat constitution and balanced constitution, which can help doctors objectively identify TCM constitution. By adding the Convolutional Attention Block Module (CBAM) to the DenseNet network, the feature extraction ability of the network can be effectively improved. We preprocess 700 voice data of 34 subjects to obtain the corresponding Mel chromatograms, and use ImageNet and AudioSet pre-trained DenseNet-CBAM model to classify them. The accuracy of our method is 82.69 %, which higher than AlexNet, ResNet and DenseNet, respectively. It can be seen that our method can improve the efficiency of constitution identification and promote the objectification of constitution identification.

Tongue segmentation is a key step to realize the intelligent tongue diagnosis of traditional Chinese medicine. The segmentation based on deep neural networks has achieved good results with a large amount of data with pixel-level annotation. But it is difficult to obtain a large amount of pixel-level labeled real data. In this work, we propose a simple and effective tongue segmentation method using synthetic data with rich pixel-level annotations. Firstly, we adopt the Fourier transform and its inverse transform to reduce the distribution difference between the target and source domains. Then we construct fold atrous spatial pyramid pooling (FASPP) to get multi-scale information and enhance the correlation among local features for tongue data with different sizes. Finally, the experimental results show the effectiveness of our approach.

Dynamic spectrum detection has attracted increasing interest in drone or drone controller detection problems. Spectrum sensing as a promising solution allows us to provide a dynamic spectrum map within the target frequency band by estimating the occupied sub-bands in a specific period. In this paper, a robust Student’s t-distribution model is built to tackle the scenario with a small number of observed samples. Then, relying on the characteristics of the statistical model, we propose an appropriate goodness-of-fit (GoF) test statistic regarding a small number of samples. Moreover, to obtain a reliable sensing, bilateral hypotheses of the test statistic are both used to make a decision. Numerical simulations show the superiority of the proposed method compared with other schemes, including the unilateral hypothesis-based GoF testing and the conventional energy detection, in a small number of sample cases.