Cracks pose a significant threat to road and building safety, making effective detection of cracks on road surfaces a focus of research both domestically and internationally. Deep learning-based methods often require extensive pixel-level annotations, posing significant labor costs. We propose a single-stage weakly supervised crack segmentation model based on multi-scale feature fusion. The model is built on a single-stage weakly supervised segmentation framework, which reduces model complexity. It utilizes a multi-scale feature fusion module (PPM) to integrate features at different scales, enhancing the ability to extract features from cracks of various sizes. The combination of the Domain Restriction Suppression (DRS) module and pixel affinity convolution is employed to optimize pseudo-pixel annotations. In addition, we propose a joint loss function to mitigate sample imbalance between crack and non-crack pixels. Compared to other two-stage weakly supervised segmentation models, our model is simpler and more effective, achieving excellent results on the Deep Crack and Crack500 datasets, surpassing most weakly supervised crack segmentation models in terms of Recall (Re), F-score (F1), and mean Intersection-over-Union (mIoU), achieving similar effects to fully supervised crack segmentation models. This demonstrates the effectiveness and robustness of our model.

Neoadjuvant chemoradiotherapy (nCRT) is the standard treatment for locally advanced rectal cancer (LARC). However, substantial inter-patient variability in response to nCRT poses a significant challenge for accurately predicting treatment outcomes based on preoperative data, thereby complicating clinical decision-making. With the rapid advancement of artificial intelligence technologies, there has been growing interest in leveraging AI for predictive modeling in cancer therapy. In this study, we propose a novel multi-modal prediction framework, FMPNet, which integrates preoperative magnetic resonance imaging (MRI) and whole-slide image (WSI) features to predict nCRT efficacy. Specifically, for WSI processing, we develop an efficient tumor cell segmentation strategy and incorporate a deep subspace clustering mechanism into the feature extraction pipeline to enhance the model's representational capacity. Comprehensive experiments demonstrate that FMPNet consistently outperforms ten other feature fusion-based prediction models on both internal test sets and external validation cohorts across multiple metrics, including accuracy, precision, recall, F1-score and ROC-AUC curves. These results not only confirm the superior performance of our model but also underscore its potential to support more accurate and personalized clinical decision-making for patients with LARC.

Automated quality assessment of cardiac ultrasound images is crucial for ensuring diagnostic accuracy and clinical decision-making reliability. Hospitals face significant challenges in efficiently screening ultrasound image quality, where manual expert review is time-consuming and subjective. However, dedicated methods for ultrasound image quality assessment remain scarce, with most adapted from natural image quality metrics that fail to capture clinically relevant diagnostic factors. In this paper, we propose a novel dual-domain framework that models both spatial anatomical features and frequency-domain spectral characteristics using specialized neural modules. Our approach incorporates cardiac-specific attention mechanisms and wavelet-based artifact detection to enable comprehensive, clinically aligned evaluation. Extensive experiments on a largescale clinical dataset demonstrate the superiority of our method, achieving 88.1 % overall accuracy, a macro-averaged F1-score of 88.1 %, and 100 % precision and recall in detecting diagnostically unacceptable images. The proposed framework is designed to meet clinical reliability standards, paving the way for safe integration into diagnostic workflows.

The incidence of ventricular tachycardia has been steadily increasing year by year. Current clinical interventions primarily involve medication and radiofrequency ablation surgery. However, due to technical limitations, precise and effective localization of ablation sites remains challenging, leading to prolonged surgery times and impacting patient outcomes. In this study, we propose using diffusion models to establish a bidirectional mapping relationship between ECG and cardiac electrophysiology, unifying the forward and inverse problem-solving processes of cardiac electrophysiological signals. Specifically, during the training phase, we add Gaussian noise to cardiac electrophysiological signals through the forward process, progressively constructing the distribution of ECG and Gaussian mixed noise. Subsequently, the reverse process is utilized to denoise the ECG and Gaussian mixed noise, learning the distribution of cardiac electrophysiological data. Ultimately, the trained diffusion model establishes a mapping relationship from ECG to cardiac electrophysiology, thus achieving the inverse problem-solving of the electrophysiological model.

Neoadjuvant chemoradiotherapy (nCRT) is the stan-dard treatment for locally advanced rectal cancer (LARC). With the development of artificial intelligence, an increasing number of studies have begun to explore its application in cancer treatment prediction. However, the prior methods exhibit considerable variability even with slight modifications to the input data, which could potentially undermine the reliability of the results. In this paper, we proposed RP-Net, a novel multi-modal fusion-based framework that combines feature information from magnetic resonance imaging (MRI) and whole slide images (WSI), establishing a relationship to map the therapeutic effectiveness of nCRT for LARC. We investigated the relationship of the tumour region and its periphery tissues, and demonstrated the validity of the proposed framework that involving 11 different combinations of modalities. The experimental results revealed that it has achieved higher prediction accuracy compared to the four intra-categories single-modal combinations and outperformed the two intra-categories multi-modal combinations. When compared to the other four inter-categories multi-modal combinations, the fusion features get accuracy of 2 % ~ 6% improvement respectively.

Liver cancer is the second leading cause of death all over the world in the 2020s’, and the incidence rate has been growing on a global scale and become a serious threat to human life. Early diagnosis of liver cancer from medical images can allow the patients to receive better treatment and achieve good outcomes. Although medical imaging approaches have made significant progress over the past decades, there are still great demands to reconstruct the network structure for the adaptation to downstream tasks. It remains a great challenge for liver tumor identification from MRI images. Recently, self-attention mechanism based transformer models can capture long-range dependencies, which make them perform well on many medical image analysis tasks. Such as Segformer and TransUNet, since lacking the translation in-variance and inductive bias of CNNs, they are still needed large-scale training to fill the gap, especially in the field of medical image analysis domain. In this study, we incorporate a novel flexible Condeathblock as a feature extractor to perform liver images analysis and propose a new analytical framework CXNet to extract discriminative multi-scale visual representations. The experiments results demonstrated our framework outperforms the state-of the-art models on three datasets, including a publicly available liver segmentation dataset as well as two in-house liver tumor classification/segmentation datasets. On the 3DIRCADb dataset, our CXNet outperforms the UNet model by 2.82%, 2.73%, and 4.46% in terms of the Jaccard similarity coefficient, Dice coefficient, and accuracy, and outperforms the TransUNet model by 8.49%, 5.35%,which 3.62%, respectively. Code is available at https://github.com/SPECTRELWF/CXNet

Cell detection and counting are critical and essential tasks for many biological and clinical studies. Traditionally, these tasks are usually performed by visual inspection, which is time consuming and prone to induce subjective bias. These make automatic cell counting and detection essential for large- scale and objective studies. Unfortunately, the hard examples such as cell blur, clutter, bleed-through and imaging noise make these tasks extremely challenging. Over the last few years, automatic cell detection and counting have evolved from earlier methods that are often based on filters to the current state-of- the-art deep learning methods. In this paper, we propose a novel efficient method for robust counting and detection task based on fully convolution networks (FCN). Our method is able to handle most of detection and counting problems from different kinds of cell datasets, and can cover most senior microscopy images, such as bright field, pathology stained material and electron. Extensive experiments on the public and private datasets demonstrate the effectiveness and reliability of our approach.

BACKGROUND: Reconstruction of neuron anatomy structure is a challenging and important task in neuroscience. However, few algorithms can automatically reconstruct the full structure well without manual assistance, making it essential to develop new methods for this task. METHODS: This paper introduces a new pipeline for reconstructing neuron anatomy structure from 3-D microscopy image stacks. This pipeline is initialized with a set of seeds that were detected by our proposed Sliding Volume Filter (SVF), given a non-circular cross-section of a neuron cell. Then, an improved open curve snake model combined with a SVF external force is applied to trace the full skeleton of the neuron cell. A radius estimation method based on a 2D sliding band filter is developed to fit the real edge of the cross-section of the neuron cell. Finally, a surface reconstruction method based on non-parallel curve networks is used to generate the neuron cell surface to finish this pipeline. RESULTS: The proposed pipeline has been evaluated using publicly available datasets. The results show that the proposed method achieves promising results in some datasets from the DIgital reconstruction of Axonal and DEndritic Morphology (DIADEM) challenge and new BigNeuron project. CONCLUSION: The new pipeline works well in neuron tracing and reconstruction. It can achieve higher efficiency, stability and robustness in neuron skeleton tracing. Furthermore, the proposed radius estimation method and applied surface reconstruction method can obtain more accurate neuron anatomy structures.

Automatic neuron reconstruction is one of the foremost challenging and important problem in the field of neuroscience. However, none of the prevalent algorithms can automatically reconstruct full anatomy structure. All of these make it is essential of developing new method for the tracing task. This paper introduced a novel seeding method for reconstructing neuron structures from 3-D microscopy images stacks. The protocol was initialized with a set of seeds which were detected by our proposed Sliding Volume Filter. And then the open curve snake was applied to the detected seeds to reconstruct the full structural of neuron cells. Results showed the proposed method exhibited excellent performance with its accuracy compared with traditional method. It is worth noting that the seeding method can clearly benefit for 3-D neuron fiber detection and reconstruction.

Hodgkin-Huxley(HH) equation is a classical model in electrophysiology and has been studied by many scholars. Applying stability theory, and taking maximal sodium conductance g̅naand potassium conductance g̅kas variables, in this study we analyze the stability and bifurcations of the model. Bifurcations are found when the variables change, and bifurcation points and boundary are calculated. When g̅nais the variable, there is only one bifurcation point and there are two points when g̅kis variable. The (g̅na, g̅k) plane is partitioned into two regions and the upper bifurcation boundary is similar to a line when both g̅naand g̅kare variables. The results gotten could be a help to control relevant diseases caused by maximal conductance anomaly.

Compared to conventional network coding, which must combine bit streams after they are decoded, the physical layer network coding (PNC) could make full use of electromagnetic (EM) waves arrived and combine bit stream at the physical layer. Because of this the system throughput could be improved. In this scheme we integrate PNC technique into wireless video transmission based on H.264 for relay channel. At the relay node the bit streams for exchanging are received and combined with different coefficients for the prioritized importance at each source node. Based on this process unequal error protection (UEP) could be realized. The proposed scheme could be outspreaded into the multi-user access model. Also a scheme of adaptively allocating the coefficient is proposed and as the simulation results depict the adaptive method could improve the quality of received image compared to conventional equal error protection (EEP) scheme. And compared to traditional video transmission the system throughput could be improved because of the use of PNC technique.

We propose a novel iterative physical layer network and channel decoding scheme based on XOR network coding and turbo code for the multiple-access relay channel. Firstly, the system model of two access nodes and one terminal with a relay between them is presented. At the relay node physical-layer network coding (PNC) is employed to combine the data information transmitted from resource, then at the terminal with the help of a check node an iterative scheme of soft information exchanging between PNC decoding and channel decoding is implemented. Meanwhile the complexity of the relay node also could be significantly reduced. And as the simulation results depict this scheme outperforms other joint decoding schemes and the complexity at the relay node is also reduced, especially in the case the channel condition between one transmitter and receiver is poor. Then multiple access nodes model is presented, also the scheme could propose good performance by contrast with other schemes only perform better in two sources case.

In this paper, by jointly considering power and delay constraints, we present an unequal error protection (UEP) approach for minimization of the receiver power consumption subject to a given quality of service, by exploiting data partition and selective hybrid automatic repeat request (HARQ). The proposed scheme assigns different error resilient tools to different priority data partitions with UEP and delay-constrained ARQ. Simulations are conducted to H.264 video over fading channels, and achieve excellent tradeoff between video delivery quality and power consumption. Simulation results demonstrate that the proposed scheme yields obvious better video quality compared with the classical unequal error protection approach and significant power saving compared with hybrid UEP (HUEP) scheme which adopts different HARQ mechanisms to different priority data in wireless video transmission.

In this paper, we present a new error robust method for H.264 video transmission. This scheme uses unequal error protection (UEP) and automatic repeat request (ARQ) protocol to provide better error-resilience performance and efficiently bandwidth utilization. In order to realize the layered video coding with transport prioritization, the proposed scheme adopts different HARQ (hybrid ARQ) schemes for the different priority-data. Simulations are conducted over the AWGN and Rayleigh channels, respectively. Simulation results demonstrate that the UEP scheme improves transmission performance and achieves better video quality than equal error protection (EEP) scheme over wireless channels under the same bandwidth.