U-Net Network Research Articles

Graph Neural Network Enhanced Dual-Branch Network for lesion segmentation in ultrasound images

Ultrasound images are widely used in the detection of breast lesions, thyroid nodules, renal tumors and other medical conditions due to their cost-effectiveness and convenience. Automatic segmentation of lesions in ultrasound images is crucial for clinical diagnosis. U-shaped networks with encoder–decoder architectures and skip connections have achieved significant success in medical image segmentation. However, challenges arise in segmentation due to the inherent blurriness of ultrasound image boundaries and the irregular shapes and volumes of lesions. Despite the capability of transformers to capture global information, we posit that graph neural network (GNN) offers a more flexible approach to capturing irregular features. GNN constructs connections for each part, leveraging global information in the process. In this study, we propose GNN-Enhanced Dual-Branch Network (GED-Net), a robust framework for lesion segmentation in ultrasound images. The Network consists of an encoder with two branches: one for extracting local features using convolutional neural network (CNN) and the other for extracting global features using GNN, along with a fusion module and a decoder. We evaluated the model on four publicly available medical ultrasound datasets including BUSI, Dataset B, DDTI and OASBUD, and demonstrated superior performance compared to most classical U-Net networks and state-of-the-art U-shaped networks. The source code is available at https://github.com/Yakiw/GED-Net.

Complex-valued multi-frequency electrical impedance tomography based on deep neural networks

The utilization of multi-frequency electrical impedance tomography (mfEIT), a non-invasive imaging technique, allows for the visualization of the conductivity distribution in biological tissues across different frequencies. However, the analysis of phase angle information within complex impedance remains a challenge, as most existing deep learning-based mfEIT algorithms are limited to real number processing. To mitigate this limitation, this study proposes a comlex reconstruction method which is inspired by the idea of combining deep learning with traditional reconstruction algorithm. It uses a spare Bayesian learning algorithm in the preprocessing stage that can perform complex arithmetic operations, and fully learns and makes use of the correlation between the real and imaginary parts to reconstruc the distribution of complex-valued conductivity in the measurement area. After that, an altered UNet network is used to further optimize the pre-reconstruction outcomes. The experimental outcomes validate the efficacy of the proposed algorithm in accurately reconstructing the complex-valued conductivity distributions of diverse biological tissues, such as potato and pig kidney, across different frequencies. Furthermore, the algorithm exhibits exceptional performance in mitigating the presence of image artifacts during the reconstruction process.

Deep learning-based automated liver contouring using a small sample of radiotherapy planning computed tomography images

IntroductionNo study has yet investigated the minimum amount of data required for deep learning-based liver contouring. Therefore, this study aimed to investigate the feasibility of automated liver contouring using limited data. MethodsRadiotherapy planning Computed tomography (CT) images were subjected to various preprocessing methods, such as denoising and windowing. Segmentation was conducted using the modified Attention U-Net and Residual U-Net networks. Two different modified networks were trained separately for different training sizes. For each architecture, the model trained with the training set size that achieved the highest dice similarity coefficient (DSC) score was selected for further evaluation. Two unseen external datasets with different distributions from the training set were also used to examine the generalizability of the proposed method. ResultsThe modified Residual U-Net and Attention U-Net networks achieved average DSCs of 97.62% and 96.48%, respectively, on the test set, using 62 training cases. The average Hausdorff distances (AHDs) for the modified Residual U-Net and Attention U-Net networks were 0.57 mm and 0.71 mm, respectively. Also, the modified Residual U-Net and Attention U-Net networks were tested on two unseen external datasets, achieving DSCs of 95.35% and 95.82% for data from another center and 95.16% and 94.93% for the AbdomenCT-1K dataset, respectively. ConclusionThis study demonstrates that deep learning models can accurately segment livers using a small training set. The method, utilizing simple preprocessing and modified network architectures, shows strong performance on unseen datasets, indicating its generalizability. Implications for practiceThis promising result suggests its potential for automated liver contouring in radiotherapy planning.

Involution Transformer Based U-Net for Landmark Detection in Ultrasound Images for Diagnosis of Infantile DDH.

The B-mode ultrasound based computer-aided diagnosis (CAD) has demonstrated its effectiveness for diagnosis of Developmental Dysplasia of the Hip (DDH) in infants, which can conduct the Graf's method by detecting landmarks in hip ultrasound images. However, it is still necessary to explore more valuable information around these landmarks to enhance feature representation for improving detection performance in the detection model. To this end, a novel Involution Transformer based U-Net (IT-UNet) network is proposed for hip landmark detection. The IT-UNet integrates the efficient involution operation into Transformer to develop an Involution Transformer module (ITM), which consists of an involution attention block and a squeeze-and-excitation involution block. The ITM can capture both the spatial-related information and long-range dependencies from hip ultrasound images to effectively improve feature representation. Moreover, an Involution Downsampling block (IDB) is developed to alleviate the issue of feature loss in the encoder modules, which combines involution and convolution for the purpose of downsampling. The experimental results on two DDH ultrasound datasets indicate that the proposed IT-UNet achieves the best landmark detection performance, indicating its potential applications.

CPP-UNet: Combined Pyramid Pooling Modules in the U-Net Network for Kidney, Tumor and Cyst Segmentation

CPP-UNet: Combined Pyramid Pooling Modules in the U-Net Network for Kidney, Tumor and Cyst Segmentation

Application of the bicharacteristic attention residual pyramid for the treatment of brain tumors

Currently, surgery remains the primary treatment for craniocerebral tumors. Before doctors perform surgeries, they need to determine the surgical plan according to the shape, location, and size of the tumor; however, various conditions of different patients make the tumor segmentation task challenging. To improve the accuracy of determining tumor shape and realizing edge segmentation, a U-shaped network combining a residual pyramid module and a dual feature attention module is proposed. The residual pyramid module can enlarge the receptive field, extract multiscale features, and fuse original information, which solves the problem caused by the feature pyramid pooling where the local information is not related to the remote information. In addition, the dual feature attention module is proposed to replace the skip connection in the original U-Net network, enrich the features, and improve the attention of the model to space and channel features with large amounts of information to be used for more accurate brain tumor segmentation. To evaluate the performance of the proposed model, experiments were conducted on the public datasets Kaggle_3M and BraTS2021. Because the model proposed in this study is applicable to two-dimensional image segmentation, it is necessary to obtain the crosscutting images of fair class in the BraTS2021 dataset in advance. Results show that the model accuracy, Jaccard similarity coefficient, Dice similarity coefficient, and false negative rate (FNR) on the Kaggle_3M dataset are 0.9395, 0.8812, 0.8958, and 0.007, respectively. The model accuracy, Jaccard similarity coefficient, Dice similarity coefficient, and FNR on the BraTS2021 dataset were 0.9375, 0.9072, 0.8981, and 0.0087, respectively. Compared with existing algorithms, all the indicators of the proposed algorithm have been improved, but the proposed model still has certain limitations and has not been applied to actual clinical trials. For specific datasets, the generalization ability of the model needs to be further improved. In the future work, the model will be further improved to address the aforementioned limitations.

Self-Supervised Contrastive Learning for Automated Segmentation of Brain Tumor MRI Images in Schizophrenia

Schizophrenic patients’ brain tumor magnetic resonance imaging (MRI) images are important references for doctors to diagnose and treat schizophrenia. However, automatic segmentation of these images is a professional and tedious task. Existing methods suffer from problems such as large model parameters, long computation time, and inadequate image processing. To achieve more accurate segmentation of brain tumors, we propose brain tumor MRI images for automatic segmentation using self-supervised contrastive learning in schizophrenia patients (BTCSSSP). First, a denoising algorithm based on progressive principal component analysis approximation and adaptive clustering is designed to process the noisy MRI images. Second, a brightness-aware image enhancement algorithm is developed to address the problems of non-uniformity, unclear boundaries, and poor spatial resolution of the MRI images. Finally, a cross-scale U-Net network with selective feature fusion attention module is designed based on self-supervised contrastive learning to achieve automatic segmentation of brain tumor MRI images. The results show that the BTCSSSP method yields higher Recall and Precision than existing methods. The maximum recall is 95%, and the image segmentation precision is 95%, thus indicating good practical applicability.

Application of Convolutional Gated Recurrent Units U-Net for Distinguishing between Retinitis Pigmentosa and Cone–Rod Dystrophy

Abstract Artificial Intelligence (AI) has gained a prominent role in the medical industry. The rapid development of the computer science field has caused AI to become a meaningful part of modern healthcare. Image-based analysis involving neural networks is a very important part of eye diagnoses. In this study, a new approach using Convolutional Gated Recurrent Units (GRU) U-Net was proposed for the classifying healthy cases and cases with retinitis pigmentosa (RP) and cone–rod dystrophy (CORD). The basis for the classification was the location of pigmentary changes within the retina and fundus autofluorescence (FAF) pattern, as the posterior pole or the periphery of the retina may be affected. The dataset, gathered in the Chair and Department of General and Pediatric Ophthalmology of Medical University in Lublin, consisted of 230 ultra-widefield pseudocolour (UWFP) and ultra-widefield FAF images, obtained using the Optos 200TX device (Optos PLC). The data were divided into three categories: healthy subjects (50 images), patients with CORD (48 images) and patients with RP (132 images). For applying deep learning classification, which rely on a large amount of data, the dataset was artificially enlarged using augmentation involving image manipulations. The final dataset contained 744 images. The proposed Convolutional GRU U-Net network was evaluated taking account of the following measures: accuracy, precision, sensitivity, specificity and F1. The proposed tool achieved high accuracy in a range of 91.00%–97.90%. The developed solution has a great potential in RP diagnoses as a supporting tool.

Two-Stage Detection Algorithm for Plum Leaf Disease and Severity Assessment Based on Deep Learning

Crop diseases significantly impact crop yields, and promoting specialized control of crop diseases is crucial for ensuring agricultural production stability. Disease identification primarily relies on human visual inspection, which is inefficient, inaccurate, and subjective. This study focused on the plum red spot (Polystigma rubrum), proposing a two-stage detection algorithm based on deep learning and assessing the severity of the disease through lesion coverage rate. The specific contributions are as follows: We utilized the object detection model YOLOv8 to strip leaves to eliminate the influence of complex backgrounds. We used an improved U-Net network to segment leaves and lesions. We combined Dice Loss with Focal Loss to address the poor training performance due to the pixel ratio imbalance between leaves and disease spots. For inconsistencies in the size and shape of leaves and lesions, we utilized ODConv and MSCA so that the model could focus on features at different scales. After verification, the accuracy rate of leaf recognition is 95.3%, and the mIoU, mPA, mPrecision, and mRecall of the leaf disease segmentation model are 90.93%, 95.21%, 95.17%, and 95.21%, respectively. This research provides an effective solution for the detection and severity assessment of plum leaf red spot disease under complex backgrounds.

Dual Attention-Based 3D U-Net Liver Segmentation Algorithm on CT Images.

The liver is a vital organ in the human body, and CT images can intuitively display its morphology. Physicians rely on liver CT images to observe its anatomical structure and areas of pathology, providing evidence for clinical diagnosis and treatment planning. To assist physicians in making accurate judgments, artificial intelligence techniques are adopted. Addressing the limitations of existing methods in liver CT image segmentation, such as weak contextual analysis and semantic information loss, we propose a novel Dual Attention-Based 3D U-Net liver segmentation algorithm on CT images. The innovations of our approach are summarized as follows: (1) We improve the 3D U-Net network by introducing residual connections to better capture multi-scale information and alleviate semantic information loss. (2) We propose the DA-Block encoder structure to enhance feature extraction capability. (3) We introduce the CBAM module into skip connections to optimize feature transmission in the encoder, reducing semantic gaps and achieving accurate liver segmentation. To validate the effectiveness of the algorithm, experiments were conducted on the LiTS dataset. The results showed that the Dice coefficient and HD95 index for liver images were 92.56% and 28.09 mm, respectively, representing an improvement of 0.84% and a reduction of 2.45 mm compared to 3D Res-UNet.

Lung X-ray image segmentation algorithm based on Multihead Self-Attention Mechanism (MSAG) optimizing Unet networks

In this paper, the Multihead Self-Attention Mechanism (MSAG) is used to optimize the Unet network for accurate segmentation of lung X-ray images. By introducing the MSAG module, the ability of the Unet network to capture global and local correlations is enhanced, which effectively improves the accuracy of the segmentation results. The introduction of the multi-head self-attention mechanism enables the network to have more powerful modelling and generalization capabilities, and can process various types of lung X-ray images stably and efficiently. The dataset is divided into training, validation and test sets according to the ratio of 4:3:3. The loss gradually converges during the training process, and the model gradually learns the data features and patterns, and the gap between them and the real labels is gradually reduced. The performance on the validation set is good and no over-fitting occurs, demonstrating the ability to generalize on unseen data. The evaluation metrics on the test set show an IoU of 0.85, a Dice of 0.92, and an Accuracy of 0.88, proving that the model can accurately extract lung features for segmentation. This study has achieved satisfactory results in the field of medical images by optimizing the network structure and introducing new techniques, which are of positive significance for improving the accuracy and efficiency of lung X-ray image segmentation.

Lung X-ray image segmentation algorithm based on Multihead Self-Attention Mechanism (MSAG) optimizing Unet networks

In this paper, the Multihead Self-Attention Mechanism (MSAG) is used to optimize the Unet network for accurate segmentation of lung X-ray images. By introducing the MSAG module, the ability of the Unet network to capture global and local correlations is enhanced, which effectively improves the accuracy of the segmentation results. The introduction of the multi-head self-attention mechanism enables the network to have more powerful modelling and generalization capabilities, and can process various types of lung X-ray images stably and efficiently. The dataset is divided into training, validation and test sets according to the ratio of 4:3:3. The loss gradually converges during the training process, and the model gradually learns the data features and patterns, and the gap between them and the real labels is gradually reduced. The performance on the validation set is good and no over-fitting occurs, demonstrating the ability to generalize on unseen data. The evaluation metrics on the test set show an IoU of 0.85, a Dice of 0.92, and an Accuracy of 0.88, proving that the model can accurately extract lung features for segmentation. This study has achieved satisfactory results in the field of medical images by optimizing the network structure and introducing new techniques, which are of positive significance for improving the accuracy and efficiency of lung X-ray image segmentation.

Research on identification of nucleus-shaped anomaly regions in space electric field

The presence of nucleus-shaped anomalous regions in the power spectrum image of the electric field VLF frequency band has been discovered in previous studies. To detect and analyze these nucleus-shaped abnormal areas and improve the recognition rate of nucleus-shaped abnormal areas, this paper proposes a new nucleus-shaped abnormal area detection model ODM_Unet (Omni-dimensional Dynamic Mobile U-net) based on U-net network. Firstly, the power spectrum image data used for training is created and labeled to form a dataset of nucleus-shaped anomalous regions; Secondly, the ODConv (Omni-dimensional Dynamic Convolution) module with embedded attention mechanism was introduced to improve the encoder, extracting nucleus-shaped anomaly region information from four dimensions and focusing on the features of different input data; An SDI (Semantics and Detail Infusion) module is introduced between the encoder and decoder to solve the problem of detail semantic loss in high-level images caused by the reduction of downsampling image size; In the decoder stage, the SCSE (Spatial and Channel Squeeze-and-Excitation) attention module is introduced to more finely adjust the feature maps output through the SDI module. The experimental results show that compared with the current popular semantic segmentation algorithms, the ODM_Unet model has the best detection performance in nucleus-shaped anomaly areas. Using this model to detect data from November 2021 to October 2022, it was found that the frequency of nucleus-shaped anomaly areas is mostly between 0 and 12.5KHz, with geographic spatial distribution ranging from 40° to 70° south and north latitudes, and magnetic latitude spatial distribution ranging from 58° to 80° south and north latitudes. This method has reference significance for detecting other types of spatial electromagnetic field disturbances.

UniverDetect: Universal landmark detection method for multidomain X-ray images

Landmark detection using X-ray imaging is vital for disease screening, treatment, and prognosis. It provides a framework for subsequent tasks, including segmentation, classification, and target detection. Existing landmark detection methods exhibit limits with the increasing number of X-ray examiners. Not only they serve specific data types, but also their ability to learn the complete semantic information from this dataset is limited, potentially affecting their widespread adoption. We address these challenges by proposing a universal landmark detection model for multidomain X-ray imaging, UniverDetect. By progressively acquiring local semantic information, global semantic insights, and anatomical structure knowledge at various levels, the model ensured that the detected landmarks were closely aligned with the ground-truth labels. UniverDetect consists of three key components. The landmark detection module (LDM) utilizes a U-net network equipped with our innovative pyramid depthwise separable (PDS) convolution module for initial landmark detection. The landmark refinement module (LRM) integrates a fine-tuning module comprising continuously extended convolution blocks. Finally, the landmark correction module (LCM) incorporates a graph convolutional network (GCN) to rectify offset errors during partial landmark detection. An inherent feature of this model is its domain- generalization capability, which enables continuous learning across diverse domains. This model can concurrently learn from eight domains, covering 118 landmarks within a diverse dataset of 5969 images. Extensive experiments on multiple datasets demonstrate that this method consistently outperforms state-of-the-art approaches. This study presents a versatile and effective solution to reduce doctors’ workload while providing precise quantitative analysis.

3D measurement of precise part in complex circumstance using line structured light and improved U-Net

Abstract Line structured light scanning is extensively utilized for the 3D measurement of precise metal parts, but the curvature surfaces and specific materials of these parts generate specular reflection, making it challenging to accurately extract the center of the laser stripe in complex circumstances. Therefore, the primary challenges are the noise separation in the laser stripe image and the extraction of the laser stripe center under complex circumstances. To solve the above problems, an improved U-Net semantic segmentation algorithm is proposed by adding an attention mechanism and modifying skip connections to the classical U-Net network structure for accurate laser stripe segmentation. Secondly, the dual smoothing method of mean smoothing and Savitzky–Golay smoothing is combined with the Hessian matrix algorithm to complete the subpixel extraction of the center point of the laser stripe. Finally, taking the blade and shaft part as the measurement objects, the experimental results demonstrate that the method can obtain more complete, smoother, and denser results than the traditional method under highly reflective surfaces, vital interference spots, and strong ambient light. The proposed method is used for the 3D measurement of the shaft parts, and the diameter measurement maximum error is 0.029 mm, verifying the feasibility of the proposed method.

Singing voice separation based on U-NET neural network and its application

Singing Voice separation is a very important method that helps people to obtain a vocal or accompaniment for further usage like music composing. In this paper, I developed a Song accompaniment separation application based on Convolutional neural network Unet model and Deep learning (DL). Basically, the U-net convolutional network is applied to the two-dimensional spectrum of a music audio in order to achieve a more precise separation effect. During this deep learning process, a significant amount of data was inserted into the model to get the best fit parameters. Then, after all, the model can automatically generate a mask that filters out the accompaniment or vocal part that users want.. This completed model is then transmitted to the terminal of a internet server to enable users to get access to the model through internet. After the test of both quantitative evaluation and subjective assessment, the application impressively achieved the Song accompaniment Separation that provides a good quality of both vocal and song.

FDiff-Fusion: Denoising diffusion fusion network based on fuzzy learning for 3D medical image segmentation

In recent years, the denoising diffusion model has achieved remarkable success in image segmentation modeling. With its powerful nonlinear modeling capabilities and superior generalization performance, denoising diffusion models have gradually been applied to medical image segmentation tasks, bringing new perspectives and methods to this field. However, existing methods overlook the uncertainty of segmentation boundaries and the fuzziness of regions, resulting in the instability and inaccuracy of the segmentation results. To solve this problem, a denoising diffusion fusion network based on fuzzy learning for 3D medical image segmentation (FDiff-Fusion) is proposed in this paper. By integrating the denoising diffusion model into the classical U-Net network, this model can effectively extract rich semantic information from input medical images, thus providing excellent pixel-level representation for medical image segmentation. In this paper, a fuzzy learning module is designed on the skip path of U-Net network because of the widespread boundary uncertainty and region blurring of medical image segmentation. The module sets several fuzzy membership functions for the input encoded features to describe the similarity degree between the feature points, and applies fuzzy rules to the fuzzy membership functions, thus enhancing the modeling ability of the model for uncertain boundaries and fuzzy regions. In addition, in order to improve the accuracy and robustness of the model segmentation results, we introduced an iterative attention feature fusion method in the test phase, which added local context information to the global context information in the attention module to fuse the prediction results of each denoising time step. Finally, to validate the effectiveness of FDiff-Fusion, we compare it with existing advanced segmentation networks on the BRATS 2020 brain tumor dataset and the BTCV abdominal multi-organ dataset. The results show that FDiff-Fusion significantly improves the Dice scores and HD95 distance on these two datasets, demonstrating its superiority in medical image segmentation tasks.

Digital Image Forgery Detection And Localization Using The Innovated U-Net

A reliable image copy–move forgery detection approach adaptable to different scenarios of tampering with color images is crucial for many applications. Different methods and solutions have been effectively proposed, but they are still subject to false positive/negative detections and cannot handle the variety of copy–move forgeries. In this paper, a machine learning model that combines ResNet 50 and U-net architectures for automatic image forgery detection in color image(s) is presented. The proposed system is inspired by the ResNet 50 architecture as an encoder and the U-Net architecture as a decoder. The encoder function implies applying convolution and normalizing for feature extraction. Conversely, the decoder functions is locating the spatial features. The decoder in the U-Net network comprises multiple decoder blocks, which are connected to corresponding encoder blocks by employing concatenate layers. A binary mask is then produced to represent the tampered regions in the image. Quantitative experimental results on two standard public datasets and a comparison with state-of-the-art methods demonstrate the effectiveness and robustness of the proposed model.

MA_W-Net-Based Dual-Output Method for Microseismic Localization in Strong Noise Environments

With the continuous depletion of conventional oil and gas reservoir resources, the beginning of exploration and development of unconventional oil and gas reservoir resources has led to the rapid development of microseismic monitoring technology. Addressing the challenges of low signal-to-noise ratio and inaccurate localization in microseismic data, we propose a new neural network MA_W-Net based on the U-Net network with the following improvements: (1) The foundational U-Net model was refined by evolving the single-channel decoder into a two-channel decoder, aimed at enhancing microseismic event localization and noise suppression capabilities. (2) The integration of attention mechanisms such as the convolutional block attention module (CBAM), coordinate attention (CA), and squeeze-and-excitation (SE) into the encoder to bolster feature extraction. We use synthetic data for evaluating the proposed method. Comparing with the normal U-net network, our accuracy in seismic recordings with a signal-to-noise ratio of −15 is improved from 78 percent to 93.5 percent, and the average error is improved from 2.60 m to 0.76 m. The results show that our method can accurately localize microseismic events and denoising processes from microseismic records with a low signal-to-noise ratio.

Development of an Underwater Detection Robot for the Structures with Pile Foundation

Addressing the challenges for detecting underwater damage to the structures with pile foundation, this study presents the design of an adsorption-operated robotic system. Initially, a prototype is proposed for an adsorption-operated robot that is equipped with an automatic movement mechanism. To account for the disturbance of the drag force in the control system, a hydrodynamic numerical model for the detection robot is established. The numerical results of the drag force are then integrated into the control system, leading to the design of a force-redeemed active disturbance rejection controller. Furthermore, a novel detection algorithm based on the image segmentation network UNet is developed for the automatic identification of damages. The performance of the controller and detection algorithm is evaluated against traditional methods. The results indicate that the proposed force-redeemed active disturbance rejection controller outperforms traditional PID and ADR controllers in terms of overshoot and adjustment time. Moreover, the developed detection algorithm exhibits superior performance in precision, F1 score, and mean intersection over union compared to the traditional algorithms of FCN, PSPnet, Deeplabv3, and UNet. Overall, this paper advances the technology of autonomous robots for underwater detection in the structures with pile foundation.