Bottleneck Layer Research Articles

DAU-Net: Dual attention-aided U-Net for segmenting tumor in breast ultrasound images.

Breast cancer remains a critical global concern, underscoring the urgent need for early detection and accurate diagnosis to improve survival rates among women. Recent developments in deep learning have shown promising potential for computer-aided detection (CAD) systems to address this challenge. In this study, a novel segmentation method based on deep learning is designed to detect tumors in breast ultrasound images. Our proposed approach combines two powerful attention mechanisms: the novel Positional Convolutional Block Attention Module (PCBAM) and Shifted Window Attention (SWA), integrated into a Residual U-Net model. The PCBAM enhances the Convolutional Block Attention Module (CBAM) by incorporating the Positional Attention Module (PAM), thereby improving the contextual information captured by CBAM and enhancing the model's ability to capture spatial relationships within local features. Additionally, we employ SWA within the bottleneck layer of the Residual U-Net to further enhance the model's performance. To evaluate our approach, we perform experiments using two widely used datasets of breast ultrasound images and the obtained results demonstrate its capability in accurately detecting tumors. Our approach achieves state-of-the-art performance with dice score of 74.23% and 78.58% on BUSI and UDIAT datasets, respectively in segmenting the breast tumor region, showcasing its potential to help with precise tumor detection. By leveraging the power of deep learning and integrating innovative attention mechanisms, our study contributes to the ongoing efforts to improve breast cancer detection and ultimately enhance women's survival rates. The source code of our work can be found here: https://github.com/AyushRoy2001/DAUNet.

LightCF-Net: A Lightweight Long-Range Context Fusion Network for Real-Time Polyp Segmentation.

Automatically segmenting polyps from colonoscopy videos is crucial for developing computer-assisted diagnostic systems for colorectal cancer. Existing automatic polyp segmentation methods often struggle to fulfill the real-time demands of clinical applications due to their substantial parameter count and computational load, especially those based on Transformer architectures. To tackle these challenges, a novel lightweight long-range context fusion network, named LightCF-Net, is proposed in this paper. This network attempts to model long-range spatial dependencies while maintaining real-time performance, to better distinguish polyps from background noise and thus improve segmentation accuracy. A novel Fusion Attention Encoder (FAEncoder) is designed in the proposed network, which integrates Large Kernel Attention (LKA) and channel attention mechanisms to extract deep representational features of polyps and unearth long-range dependencies. Furthermore, a newly designed Visual Attention Mamba module (VAM) is added to the skip connections, modeling long-range context dependencies in the encoder-extracted features and reducing background noise interference through the attention mechanism. Finally, a Pyramid Split Attention module (PSA) is used in the bottleneck layer to extract richer multi-scale contextual features. The proposed method was thoroughly evaluated on four renowned polyp segmentation datasets: Kvasir-SEG, CVC-ClinicDB, BKAI-IGH, and ETIS. Experimental findings demonstrate that the proposed method delivers higher segmentation accuracy in less time, consistently outperforming the most advanced lightweight polyp segmentation networks.

SARFNet: Selective Layer and Axial Receptive Field Network for Multimodal Brain Tumor Segmentation

Efficient magnetic resonance imaging (MRI) segmentation, which is helpful for treatment planning, is essential for identifying brain tumors from detailed images. In recent years, various convolutional neural network (CNN) structures have been introduced for brain tumor segmentation tasks and have performed well. However, the downsampling blocks of most existing methods are typically used only for processing the variation in image sizes and lack sufficient capacity for further extraction features. We, therefore, propose SARFNet, a method based on UNet architecture, which consists of the proposed SLiRF module and advanced AAM module. The SLiRF downsampling module can extract feature information and prevent the loss of important information while reducing the image size. The AAM block, incorporated into the bottleneck layer, captures more contextual information. The Channel Attention Module (CAM) is introduced into skip connections to enhance the connections between channel features to improve accuracy and produce better feature expression. Ultimately, deep supervision is utilized in the decoder layer to avoid vanishing gradients and generate better feature representations. Many experiments were performed to validate the effectiveness of our model on the BraTS2018 dataset. SARFNet achieved Dice coefficient scores of 90.40, 85.54, and 82.15 for the whole tumor (WT), tumor core (TC), and enhancing tumor (ET), respectively. The results show that the proposed model achieves state-of-the-art performance compared with twelve or more benchmarks.

Denoising and Baseline Correction Methods for Raman Spectroscopy Based on Convolutional Autoencoder: A Unified Solution.

Preprocessing plays a key role in Raman spectral analysis. However, classical preprocessing algorithms often have issues with reducing Raman peak intensities and changing the peak shape when processing spectra. This paper introduces a unified solution for preprocessing based on a convolutional autoencoder to enhance Raman spectroscopy data. One is a denoising algorithm that uses a convolutional denoising autoencoder (CDAE model), and the other is a baseline correction algorithm based on a convolutional autoencoder (CAE+ model). The CDAE model incorporates two additional convolutional layers in its bottleneck layer for enhanced noise reduction. The CAE+ model not only adds convolutional layers at the bottleneck but also includes a comparison function after the decoding for effective baseline correction. The proposed models were validated using both simulated spectra and experimental spectra measured with a Raman spectrometer system. Comparing their performance with that of traditional signal processing techniques, the results of the CDAE-CAE+ model show improvements in noise reduction and Raman peak preservation.

Improving Axial-Attention Network via Cross-Channel Weight Sharing

In recent years,Hypercomplex-inspired neural networks improved deep CNN architectures due to their ability to share weights across input channels and thus improve cohesivenessof representations within the layers.The work described herein studies the effect of replacing existing layers inan Axial Attention ResNet with their quaternion variants that use cross-channel weight sharingto assess the effect on image classification. We expect the quaternion enhancements to produce improved feature maps with more interlinked representations.We experiment with the stem of the network, the bottleneck layer, and thefully connected backend, by replacing them with quaternionversions.These modifications lead to novel architectures which yield improved accuracyperformance on the ImageNet300k classification dataset. Our baseline networks for comparison were the original real-valued ResNet, the original quaternion-valuedResNet, and the Axial Attention ResNet.Since improvement was observed regardless of which part of the network was modified,there is a promise that this technique may be generally useful in improvingclassification accuracy for a large classof networks.

An Integrated Multitasking Intelligent Bearing Fault Diagnosis Scheme Based on Representation Learning Under Imbalanced Sample Condition.

Accurate bearing fault diagnosis is of great significance of the safety and reliability of rotary mechanical system. In practice, the sample proportion between faulty data and healthy data in rotating mechanical system is imbalanced. Furthermore, there are commonalities between the bearing fault detection, classification, and identification tasks. Based on these observations, this article proposes a novel integrated multitasking intelligent bearing fault diagnosis scheme with the aid of representation learning under imbalanced sample condition, which realizes bearing fault detection, classification, and unknown fault identification. Specifically, in the unsupervised condition, a bearing fault detection approach based on modified denoising autoencoder (DAE) with self-attention mechanism for bottleneck layer (MDAE-SAMB) is proposed in the integrated scheme, which only uses the healthy data for training. The self-attention mechanism is introduced into the neurons in the bottleneck layer, which can assign different weights to the neurons in the bottleneck layer. Moreover, the transfer learning based on representation learning is proposed for few-shot fault classification. Only a few fault samples are used for offline training, and high-accuracy online bearing fault classification is achieved. Finally, according to the known fault data, the unknown bearing faults can be effectively identified. A bearing dataset generated by rotor dynamics experiment rig (RDER) and a public bearing dataset demonstrates the applicability of the proposed integrated fault diagnosis scheme.

PAMSNet: A medical image segmentation network based on spatial pyramid and attention mechanism

The image segmentation of diseases can help clinical diagnosis and treatment in medical image analysis. Due to the complexity of lesion features (e.g., size, location, and morphology) and the high similarity between the background and the target area in medical images, semantic features are difficult to extract completely. To tackle these problems, we propose a novel medical image segmentation network PAMSNet based on the spatial pyramid and attention mechanism. By using efficient pyramid attention and channel spatial attention modules, the proposed method fuses the extracted multi-scale spatial information with the local features extracted by the encoder to supplement the image details. In addition, the Spatial Pyramid-Coordinate Attention (SPCA) module is introduced in the bottleneck layer to obtain larger receptive field information and enhance feature extraction. We conducted qualitative and quantitative evaluations on four public datasets, including ISIC2018, Lung segmentation, Kvasir-SEG, and ISLES2022. The segmentation accuracy of DSC was 87.86%, 98.18%, 82.43%, and 87.37%, respectively. The ablation study of each part of PAMSNet proves the validity of each component, and the comparison with state-of-the-art methods on different indicators proves the predominance of the network.

Micro LED defect detection with self-attention mechanism-based neural network

We propose a method utilizing a YOLO detector for the precise localization of defective chips and the identification of defect types within multi-scale multi-target images. To address the challenge of optimizing training costs and enhancing model generalization, we introduce an end-to-end deep neural network, CM-YOLOv5, specifically designed for chip detection. We incorporate a novel bottleneck layer, MA-CSP, in conjunction with Multi-Head Self-Attention mechanism (MHSA). Additionally, we propose a class-balanced loss function (CB-BCE Loss) to tackle the issue of uneven distribution of defective samples in the Micro LED dataset. To further enhance convergence speed and detection precision, we introduce the SIoU Loss combined with Meta-AconC. Our experimental results, conducted on the Micro LED dataset, demonstrate notable improvements with CM-YOLOv5 over the basic YOLOv5 algorithm. Specifically, CM-YOLOv5 exhibits a 3.8 % increase in mean average precision and a 3.7 % improvement in precision, surpassing current mainstream object detection algorithms, including YOLOR, YOLOX, and YOLOv6, etc., in terms of general evaluation metrics. Finally, upon deploying our proposed algorithm on the edge device NVIDIA Jetson Xavier NX, CM-YOLOv5 demonstrates commendable speed and detection performance in embedded scenarios.

Lightweight Small-Tailed Han Sheep Facial Recognition Based on Improved SSD Algorithm

We propose a lightweight detection algorithm based on the Single Shot MultiBox Detector (SSD) algorithm in order to facilitate sheep management and to realize sheep facial identification, and we take the self-constructed dataset as the research object. First, the SSD replaces the VGG16 backbone network with MobileNetv3, a lightweight neural network, to create a hybrid model that is much smaller. Second, the ECA attention mechanism is incorporated into the backend of the 72 × 160 bottleneck layer. Finally, the SmoothL1 loss function is substituted with the BalancedL1 loss function. The optimized model’s size decreases significantly from the original SSD’s 132 MB to just 22.4 MB. It achieves a mean average precision of 83.47% and maintains an average frame rate of 68.53 frames per second. Compared to the basic SSD model, the mean average precision has increased by 3.25 percentage points, the model size has decreased by 109.6 MB, and the detection speed has improved by 9.55 frames per second. In comparative experiments using the same dataset with different object detection models, the proposed model outperforms the SSD, Faster R-CNN, Retinanet, and CenterNet in terms of mean average precision, with improvements of 3.25 percentage points, 4.71 percentage points, 2.38 percentage points, and 8.13 percentage points, respectively. The detection speed has shown significant improvements, increasing by 9.55, 58.55, 53.1, and 12.37 frames per second, respectively. The improved model presented in this paper significantly reduces the model’s size and computational requirements while maintaining an excellent performance. This provides a valuable reference for the digitalization of animal husbandry and livestock farming.

Fault diagnosis of rolling bearings based on dynamic convolution and dual-channel feature fusion under variable working conditions

Addressing the challenge of inconsistent data feature distribution and the difficulty of fault diagnosis in rolling bearings operating under variable conditions, a novel approach is proposed for bearings fault diagnosis. Dynamic convolution and dual-channel feature fusion are utilized in our method. In the shallow network layer, we employ a dual-channel convolutional structure, combining a large convolutional group with a small convolutional group to enhance the extraction of high-low frequency fault information from images. The improved GhostNetV2 bottleneck layer was used in the deeper layer of the network to obtain more beneficial features through the dynamic convolution and attention mechanism. Finally, fault classification and evaluation under variable working conditions was performed on the Case Western Reserve University and Drivetrain Dynamic Simulator (DDS) datasets. Our results showed that the methods and model used in this study can effectively handle the precision fault detection across various operational scenarios.

HM_ADET: a hybrid model for automatic detection of eyelid tumors based on photographic images

BackgroundThe accurate detection of eyelid tumors is essential for effective treatment, but it can be challenging due to small and unevenly distributed lesions surrounded by irrelevant noise. Moreover, early symptoms of eyelid tumors are atypical, and some categories of eyelid tumors exhibit similar color and texture features, making it difficult to distinguish between benign and malignant eyelid tumors, particularly for ophthalmologists with limited clinical experience.MethodsWe propose a hybrid model, HM_ADET, for automatic detection of eyelid tumors, including YOLOv7_CNFG to locate eyelid tumors and vision transformer (ViT) to classify benign and malignant eyelid tumors. First, the ConvNeXt module with an inverted bottleneck layer in the backbone of YOLOv7_CNFG is employed to prevent information loss of small eyelid tumors. Then, the flexible rectified linear unit (FReLU) is applied to capture multi-scale features such as texture, edge, and shape, thereby improving the localization accuracy of eyelid tumors. In addition, considering the geometric center and area difference between the predicted box (PB) and the ground truth box (GT), the GIoU_loss was utilized to handle cases of eyelid tumors with varying shapes and irregular boundaries. Finally, the multi-head attention (MHA) module is applied in ViT to extract discriminative features of eyelid tumors for benign and malignant classification.ResultsExperimental results demonstrate that the HM_ADET model achieves excellent performance in the detection of eyelid tumors. In specific, YOLOv7_CNFG outperforms YOLOv7, with AP increasing from 0.763 to 0.893 on the internal test set and from 0.647 to 0.765 on the external test set. ViT achieves AUCs of 0.945 (95% CI 0.894-0.981) and 0.915 (95% CI 0.860-0.955) for the classification of benign and malignant tumors on the internal and external test sets, respectively.ConclusionsOur study provides a promising strategy for the automatic diagnosis of eyelid tumors, which could potentially improve patient outcomes and reduce healthcare costs.

Spectral encoder to extract the efficient features of Raman spectra for reliable and precise quantitative analysis

Raman spectroscopy has become a powerful analytical tool highly demanded in many applications such as microorganism sample analysis, food quality control, environmental science, and pharmaceutical analysis, owing to its non-invasiveness, simplicity, rapidity and ease of use. Among them, quantitative research using Raman spectroscopy is a crucial application field of spectral analysis. However, the entire process of quantitative modeling largely relies on the extraction of effective spectral features, particularly for measurements on complex samples or in environments with poor spectral signal quality. In this paper, we propose a method of utilizing a spectral encoder to extract effective spectral features, which can significantly enhance the reliability and precision of quantitative analysis. We built a latent encoded feature regression model; in the process of utilizing the autoencoder for reconstructing the spectrometer output, the latent feature obtained from the intermediate bottleneck layer is extracted. Then, these latent features are fed into a deep regression model for component concentration prediction. Through detailed ablation and comparative experiments, our proposed model demonstrates superior performance to common methods on single-component and multi-component mixture datasets, remarkably improving regression precision while without needing user-selected parameters and eliminating the interference of irrelevant and redundant information. Furthermore, in-depth analysis reveals that latent encoded feature possesses strong nonlinear feature representation capabilities, low computational costs, wide adaptability, and robustness against noise interference. This highlights its effectiveness in spectral regression tasks and indicates its potential in other application fields. Sufficient experimental results show that our proposed method provides a novel and effective feature extraction approach for spectral analysis, which is simple, suitable for various methods, and can meet the measurement needs of different real-world scenarios.

MCCGAN: An All-In-One Image Restoration Under Adverse Conditions Using Multidomain Contextual Conditional Gan

Clear images are crucial for the optimal performance of various high-level vision-based tasks. However, some inevitable causes, such as bad weather and underwater conditions degrade scene visibility. The tiny particles present in the air absorb and scatter light, causing severe attenuation that results in unclear, low-brightness, and poor-contrast images. Several techniques have been introduced to restore the degradation. However, no model exists to date that can restore multiple degradations using a single model. Therefore, to improve the scene visibility, a unified model called a Multidomain Contextual Conditional Generative Adversarial Network (MCCGAN) is designed, which uses the same parameters across the domains to restore multiple degradations such as fog, haze, rain streaks, snowflakes, smoke, shadows, underwater, and muddy underwater. The proposed model has a novel addition of multiple [Formula: see text] convolutional context encoding bottleneck layers between a simple lightweight eight-block encoder and decoder with skip connections which learns the context of each input domain thoroughly, thus generating better-restored images. The MCCGAN is qualitatively and quantitatively compared to various state-of-the-art image-to-image translation models and tested on a few real unseen image domains such as smog, dust, and lightning, and the obtained results successfully improved scene visibility, proving the generalizability of MCCGAN. Moreover, the MS-COCO 2017 validation dataset is used for comparing the performance of object detection, instance segmentation, and image captioning on (1) weather-degraded images, (2) restored images by MCCGAN, and (3) ground truth images, and the results demonstrated the success of our model. An ablation study is also carried out to check the significance of the discriminator, skip connections, and bottleneck layers in MCCGAN, and the analysis suggests that MCCGAN performs better by adding a discriminator, skip connections, and four bottleneck layers in the generator architecture.

Deep Learning-Based Ensemble Approach for Autonomous Object Manipulation with an Anthropomorphic Soft Robot Hand

Autonomous object manipulation is a challenging task in robotics because it requires an essential understanding of the object’s parameters such as position, 3D shape, grasping (i.e., touching) areas, and orientation. This work presents an autonomous object manipulation system using an anthropomorphic soft robot hand with deep learning (DL) vision intelligence for object detection, 3D shape reconstruction, and object grasping area generation. Object detection is performed using Faster-RCNN and an RGB-D sensor to produce a partial depth view of the objects randomly located in the working space. Three-dimensional object shape reconstruction is performed using U-Net based on 3D convolutions with bottle-neck layers and skip connections generating a complete 3D shape of the object from the sensed single-depth view. Then, the grasping position and orientation are computed based on the reconstructed 3D object information (e.g., object shape and size) using U-Net based on 3D convolutions and Principal Component Analysis (PCA), respectively. The proposed autonomous object manipulation system is evaluated by grasping and relocating twelve objects not included in the training database, achieving an average of 95% successful object grasping and 93% object relocations.

Network depth affects inference of gene sets from bacterial transcriptomes using denoising autoencoders.

The increasing number of publicly available bacterial gene expression data sets provides an unprecedented resource for the study of gene regulation in diverse conditions, but emphasizes the need for self-supervised methods for the automated generation of new hypotheses. One approach for inferring coordinated regulation from bacterial expression data is through neural networks known as denoising autoencoders (DAEs) which encode large datasets in a reduced bottleneck layer. We have generalized this application of DAEs to include deep networks and explore the effects of network architecture on gene set inference using deep learning. We developed a DAE-based pipeline to extract gene sets from transcriptomic data in Escherichia coli, validate our method by comparing inferred gene sets with known pathways, and have used this pipeline to explore how the choice of network architecture impacts gene set recovery. We find that increasing network depth leads the DAEs to explain gene expression in terms of fewer, more concisely defined gene sets, and that adjusting the width results in a tradeoff between generalizability and biological inference. Finally, leveraging our understanding of the impact of DAE architecture, we apply our pipeline to an independent uropathogenic E.coli dataset to identify genes uniquely induced during human colonization. https://github.com/BarquistLab/DAE_architecture_exploration.

Lightweight and fast visual detection method for 3C assembly

In the context of 3C assembly scenarios, characterized by numerous semi-flexible, heterogeneous, and small slender targets, traditional target detection algorithms face significant challenges such as low accuracy, weak generalization, large model sizes, and slow inference speeds. To address these issues, this study introduces an enhanced method based on the YOLOv5 model, named YOLOv5-GTB. This method integrates Bidirectional Feature Pyramid Network (BiFPN), Ghost lightweight convolution, Vision Transformer (ViT), and adaptive activation function technologies, aiming to improve the accuracy and speed of target detection. Our approach utilizes Ghost convolution to construct a Ghost bottleneck layer, optimizing the feature extraction network while significantly reducing computational costs and enhancing the convolutional neural network’s feature extraction capabilities. Additionally, the Cross-Stage Partial Network (CSPNet) architecture is employed to effectively segment the data flow of the input feature map layer, thereby improving the efficiency of gradient processing. Furthermore, we introduce a fusion structure of CNN and Transformer, leveraging the strengths of convolutional neural networks in local feature extraction and the ViT’s capability in long-range information capture, thereby further enhancing the overall network’s feature extraction performance. Regarding feature fusion, considering the limitations of traditional top-down unidirectional information flow in effectively merging features with both location and semantic information, BiFPN is incorporated into the YOLOv5-GTB. This enhances the fusion of feature layers extracted from both the end of the backbone network and the first module’s fused feature layer, thus improving detection accuracy. Ablation and comparative experiments conducted on the 3C assembly scenario target detection dataset demonstrate the significant advantages of the YOLOv5-GTB model in terms of accuracy and speed. Ultimately, the application of this model to the 3C assembly platform successfully achieves rapid and accurate target recognition in this scenario.

CADS: A Self-supervised Learner via Cross-modal Alignment and Deep Self-distillation for CT Volume Segmentation.

Self-supervised learning (SSL) has long had great success in advancing the field of annotation-efficient learning. However, when applied to CT volume segmentation, most SSL methods suffer from two limitations, including rarely using the information acquired by different imaging modalities and providing supervision only to the bottleneck encoder layer. To address both limitations, we design a pretext task to align the information in each 3D CT volume and the corresponding 2D generated X-ray image and extend self-distillation to deep self-distillation. Thus, we propose a self-supervised learner based on Cross-modal Alignment and Deep Self-distillation (CADS) to improve the encoder's ability to characterize CT volumes. The cross-modal alignment is a more challenging pretext task that forces the encoder to learn better image representation ability. Deep self-distillation provides supervision to not only the bottleneck layer but also shallow layers, thus boosting the abilities of both. Comparative experiments show that, during pre-training, our CADS has lower computational complexity and GPU memory cost than competing SSL methods. Based on the pre-trained encoder, we construct PVT-UNet for 3D CT volume segmentation. Our results on seven downstream tasks indicate that PVT-UNet outperforms state-of-the-art SSL methods like MOCOv3 and DiRA, as well as prevalent medical image segmentation methods like nnUNet and CoTr. Code and pre-trained weight will be available at https://github.com/yeerwen/CADS.

DopUS-Net: Quality-Aware Robotic Ultrasound Imaging Based on Doppler Signal

Medical ultrasound (US) is widely used to evaluate and stage vascular diseases, in particular for the preliminary screening program, due to the advantage of being radiation-free. However, automatic segmentation of small tubular structures (e.g., the ulnar artery) from cross-sectional US images is still challenging. To address this challenge, this paper proposes the DopUS-Net and a vessel re-identification module that leverage the Doppler effect to enhance the final segmentation result. Firstly, the DopUS-Net combines the Doppler images with B-mode images to increase the segmentation accuracy and robustness of small blood vessels. It incorporates two encoders to exploit the maximum potential of the Doppler signal and recurrent neural network modules to preserve sequential information. Input to the first encoder is a two-channel duplex image representing the combination of the grey-scale Doppler and B-mode images to ensure anatomical spatial correctness. The second encoder operates on the pure Doppler images to provide a region proposal. Secondly, benefiting from the Doppler signal, this work first introduces an online artery re-identification module to qualitatively evaluate the real-time segmentation results and automatically optimize the probe pose for enhanced Doppler images. This quality-aware module enables the closed-loop control of robotic screening to further improve the confidence and robustness of image segmentation. The experimental results demonstrate that the proposed approach with the re-identification process can significantly improve the accuracy and robustness of the segmentation results (Dice score: from <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.54$</tex-math> </inline-formula> to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.86$</tex-math> </inline-formula> ; intersection over union: from <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.47$</tex-math> </inline-formula> to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.78$</tex-math> </inline-formula> ). <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —The Doppler signal is important for the diagnosis of vascular disease, e.g., peripheral arterial disease, in clinical practices, nevertheless it is not of similar significance for state-of-the-art robotic ultrasound (US) examination systems yet. This paper explores various neural network structures to effectively extract the blood vessels from US images by incorporating the Doppler signal into the segmentation process. The final DopUS structure with two encoders extracting differentiated information from two different inputs and fusing the latent feature representations in the bottleneck layer can also inspire other tasks like multi-senor fusion. In addition, this work developed a Doppler-based tracker to assess the quality of the segmentation results in real-time. The assessment is subsequently used for a quality-aware module that enables closed-loop control of the robotic screening. Preliminary physical experiments suggest that the quality-aware robotic screening system can improve the confidence and robustness of autonomous US examination results. In the future, the Doppler signal could also be used to support clinical diagnosis. We believe the proposed quality-aware autonomous screening system is important for the development of large-scale robotic US screening programs. It will not only benefit the examination of limb arteries but also other vascular structures, e.g., carotid or aorta.

Vibration-based novelty detection using autoencoders: application to KW51 bridge

In recent years an increasing attention has been given to the application of Artificial Intelligence techniques within the context of Structural Health Monitoring (SHM) of Civil Engineering structures. Multiple advantages compared to traditional techniques based on Operational Modal Analysis (OMA) are achieved, such as a relatively low computational efforts and possible insensitivity to environmental and operational variability (EOV). The present paper proposes an autoencoder-based approach to automatically handle EOVs and, at the same time, to detect structural changes. The proposed procedure is based on the use of an autoencoder (AE) model to reconstruct the measurement data collected during continuous dynamic monitoring. In more details, an AE network is trained using the time series simultaneously collected by all the available channels during a reference period, in which the structure is supposed to be in healthy condition under normal EOV. Through the training procedure, the internal bottleneck layer of AE is supposed to learn how the variation of measured data is affected by the normal EOV. Subsequently, the trained AE is used to reconstruct the data collected from unknown scenarios, providing a reconstruction error (between the measured and the reconstructed signals) that increases as soon as the monitored system departs from its healthy condition. The application of the AE-based procedure is exemplified to the benchmark KW51 bridge (a steel bowstring railway bridge in Leuven, Belgium), showing that the AE network, trained simultaneously with all the available channels, is capable of detecting the structural changes due to a retrofitting performed on the bridge, under changing environmental and operational conditions.

Research on the construction of marine creatures classification and identification model based on ResNet50

There are many kinds of Marine organisms and their biological forms differ greatly, so it is difficult to guarantee the accuracy of artificial species identification, which brings great challenges to the work of Marine species identification. In this paper, we propose a recognition method of Marine biological image classification using residual neural network, redefining convolution layer and using batch regularization to avoid gradient parameter disorder. The bottleneck layer is realized by the residual connection in the neural network, and the residual network ResNet50 is constructed by the transfer learning method. The classification training was conducted on 19 common Marine animal data sets, and the experimental results showed that the recognition accuracy of ResNet50 reached about 90%. Compared with the traditional convolutional neural network VGG19, the results showed that the recognition efficiency of ResNet50 was better, thus verifying the effectiveness of the Marine animal classification and recognition model proposed in this paper.