Decoding Path Research Articles

Revealing in vivo broiler chicken growth state: Integrating CT imaging and deep learning for non-invasive reproductive phenotypic measurement

The measurement of body dimensions in broiler chickens has long been of great interest to the poultry industry, especially regarding reproductive phenotypes closely linked to breeding traits and the growth status of roosters. Existing measurement methods suffer from subjectivity, destructiveness, and various limitations. Therefore, we propose a novel method for the in vivo measurement of broiler chicken reproductive phenotype based on a three-dimensional medical imaging segmentation network. Leveraging our proposed FDTR medical imaging segmentation algorithm, this method achieves robust segmentation performance on a broiler chicken dataset. However, due to the morphological diversity and extreme class imbalance of broiler chicken reproductive phenotypes, we introduce a meta-learning paradigm and design a specialized decoder path. Ultimately, we establish relationships between the predicted results obtained from the segmentation network and the reproductive phenotypes to achieve phenotypic measurement. To evaluate this method, we collected and constructed the YF2023 broiler chicken reproductive phenotype dataset. Extensive cross-validation experiments demonstrate that the proposed method effectively measures the reproductive phenotypes of live roosters using computed tomography (CT) medical imaging. The proposed segmentation model achieves a mean fold dice similarity coefficient (DSC) of 0.6353, a mean fold Hausdorff distance (HD) of 11.84 mm, and a mean fold average volume distance (AVD) of 1.72 mm3. In addition, we evaluated our model’s generalization performance on unseen data, where the model achieved a DSC of 0.6825, an HD of 7.58 mm, and an AVD of 1.40 mm3. The predicted results exhibit a mean correlation coefficient of 66.18% with the weights of reproductive phenotype. This validates the enormous potential of the method in broiler chicken breeding and provides new insights and solutions for the accurate segmentation of broiler chicken reproductive phenotype. The code is available at https://github.com/Github-XKou/MetaFDTR.

A path splitting and pruning strategy on list decoder for PAC codes

AbstractRecently, Arıkan proposed a polarization‐adjusted convolutional (PAC) codes, demonstrating their superior error correction performance over polar codes at short block lengths. It was confirmed that PAC codes approached the optimal performance achievable with limited code length. This paper proposes a novel low‐complexity list decoding algorithm for PAC codes, incorporating path splitting and pruning strategies based on a set of highly reliable information bits. Simulation results reveal that the proposed algorithm significantly reduces sorting complexity and average list size, all while incurring negligible performance loss. Unlike previous pruning algorithms designed for polar codes, the proposed strategy eliminates the need to individually assess the reliability of decoding paths in each decoding process. Instead, the algorithm minimizes redundant decoding paths through a high‐reliability information bit set, constructed using Monte Carlo experiments.

Spatial attention guided cGAN for improved salient object detection

Recent research shows that Conditional Generative Adversarial Networks (cGANs) are effective for Salient Object Detection (SOD), a challenging computer vision task that mimics the way human vision focuses on important parts of an image. However, implementing cGANs for this task has presented several complexities, including instability during training with skip connections, weak generators, and difficulty in capturing context information for challenging images. These challenges are particularly evident when dealing with input images containing small salient objects against complex backgrounds, underscoring the need for careful design and tuning of cGANs to ensure accurate segmentation and detection of salient objects. To address these issues, we propose an innovative method for SOD using a cGAN framework. Our method utilizes encoder-decoder framework as the generator component for cGAN, enhancing the feature extraction process and facilitating accurate segmentation of the salient objects. We incorporate Wasserstein-1 distance within the cGAN training process to improve the accuracy of finding the salient objects and stabilize the training process. Additionally, our enhanced model efficiently captures intricate saliency cues by leveraging the spatial attention gate with global average pooling and regularization. The introduction of global average pooling layers in the encoder and decoder paths enhances the network's global perception and fine-grained detail capture, while the channel attention mechanism, facilitated by dense layers, dynamically modulates feature maps to amplify saliency cues. The generated saliency maps are evaluated by the discriminator for authenticity and gives feedback to enhance the generator's ability to generate high-resolution saliency maps. By iteratively training the discriminator and generator networks, the model achieves improved results in finding the salient object. We trained and validated our model using large-scale benchmark datasets commonly used for salient object detection, namely DUTS, ECSSD, and DUT-OMRON. Our approach was evaluated using standard performance metrics on these datasets. Precision, recall, MAE and Fβ score metrics are used to evaluate performance. Our method achieved the lowest MAE values: 0.0292 on the ECSSD dataset, 0.033 on the DUTS-TE dataset, and 0.0439 on the challenging and complex DUT-OMRON dataset, compared to other state-of-the-art methods. Our proposed method demonstrates significant improvements in salient object detection, highlighting its potential benefits for real-life applications.

A shape-supervised feature fusion U-Net for tubular structure segmentation

Accurate segmentation of tubular structures, such as blood vessels and bile ducts, is pivotal for clinical diagnosis and subsequent treatment. However, challenges arise from their unique structure and morphology. These structures are thin and elongated, and often exhibit complex branching patterns, making precise delineation difficult. Additionally, the intricate and often noisy backgrounds in medical images complicate the extraction of detailed and shape features. These challenges collectively hinder the effectiveness of automated segmentation methods. To address these issues, this paper introduces the shape-supervised feature fusion U-Net (SFU-Net), a novel approach that extends the traditional U-Net architecture. Our network incorporates a feature fusion module that effectively combines different semantic information in both the encoder and decoder paths, facilitating the capture of detailed geometric representations of tubular structures. Furthermore, a novel distance-based tubularity loss is designed to capture boundary information and supervise segmentation integrity by imposing shape constraints. We conducted comprehensive experiments on three tubular structure segmentation tasks in medical imaging, including the dilated biliary tree, hepatic vessel, and pulmonary artery. The experimental results illustrate that the proposed SFU-Net achieves the best performance in terms of accuracy and completeness, with Dice scores of 76.14%, 80.42%, and 83.79%, and HD95 values of 8.59 mm, 6.43 mm, and 4.31 mm across three datasets, respectively. Compared to traditional U-Net-based segmentation methods and other tubular segmentation approaches, our method exhibits superior performance, underscoring its effectiveness in medical image segmentation with tubular structures.

Flow field prediction in bed configurations: A parametric spatio-temporal convolutional autoencoder approach

The use of deep learning methods for modeling fluid flow has drawn a lot of attention in the past few years. Here we present a data-driven reduced-order model (ROM) for predicting flow fields in a bed configuration of hot particles. The ROM consists of a parametric spatio-temporal convolutional autoencoder. The neural network architecture comprises two main components. The first part resolves the spatial and temporal dependencies present in the input sequence, while the second part of the architecture is responsible for predicting the solution at the subsequent timestep based on the information gathered from the preceding part. We also propose the utilization of a post-processing non-trainable output layer following the decoding path to incorporate the physical knowledge, e.g. no-slip condition, into the prediction. The ROM is evaluated by comparing its predicted solution with the high-fidelity counterpart. In addition, proper orthogonal decomposition (POD) is employed to systematically analyze and compare the dominant structures present in both sets of solutions. The assessment of the ROM for a bed configuration with variable particle temperature showed accurate results at a fraction of the computational cost required by traditional numerical simulation methods.

Segmentation and visualization of Retinal Detachment lesions through Retinal fundus images

Medical image segmentation is a prominent task in medical image processing and a fundamental aspect of medical image analysis. However, accurate segmentation is exceedingly challenging due to the variations in size, shape, and location of the lesions. U-Net is well-linked in the domain of medical image segmentation, but it fails to fully exploit the advantageous characteristics of the channel or leverage contextual information. This article presents a novel, improved MultiResU-Net framework called DCA-MultiResU-Net to minimize the loss of crucial features and also to enhance the efficiency of end-to-end image segmentation by utilizing the full advantages of U-Net, Res2Net, Channel attention mechanism, Dilated convolution layers, and MultiResUNet models. In contrast to the typical convolutions of U-Net in the encoder path, the dilated convolution modules are inserted into the MultiRes blocks to extract and concatenate the multi-scale features. Furthermore, a channel attention mechanism is employed in the bottom portion of the “U” shaped network and the decoder path to fuse the information from receptive fields of varying sizes. Afterward, the virtual reality visualization technology was implemented for the 3-D visualization of segmented RD lesions. The proposed segmentation model was evaluated on four online databases: Retinal Image Bank, RIADD, Kaggle, and Cataract Image Dataset or GitHub. The performance outcomes of this work were measured using standard evaluation indexes such as Dice similarity coefficient, Intersection over the union, F1-score, and Loss function score of 90.76%, 88.42%, 93.28%, and 0.1641, respectively. The experimental findings indicate that the DA-MultiResU-Net framework achieved superior performance and generalization capabilities while utilizing 6.76% fewer parameters than the U-Net model.

Chaotic Phase Modulation Direct-Sequence Spread Spectrum-Assisted Adaptive Serial Cancellation List Decoding Method for Underwater Acoustic Communication

Addressing the challenges of high decoding latency, reduced spectral efficiency, and substantial storage requirements in a Cyclic Redundancy Check (CRC)Aided Successive Cancellation List (CA-SCL) polar decoder, this paper proposes a chaotic phase modulation direct-sequence spread spectrum (CPMDSSS)-assisted adaptive serial cancellation list decoding method for underwater acoustic communication. The method involves segmenting the information bits to be transmitted, computing CRC for each segment, and mapping all CRCs to a CPMDSSS, which is then modulated onto the pilot subcarriers of underwater acoustic orthogonal frequency-division multiplexing (OFDM) to increase spectral efficiency. At the receiving end, CRCs are obtained by demodulating the CPMDSSS to verify the segmented information and adaptively select the number of decoding paths. The theoretical analysis and simulation results demonstrate that compared to CA-SCL, the proposed method effectively reduces the required storage units and improves spectral efficiency, with an average reduction of approximately 80% in the decoding paths. Sea trials further indicate that the proposed method reduces the average decoding paths by approximately 71% and decreases the average decoding delay by approximately 64% compared to CA-SCL.

Cross-patch feature interactive net with edge refinement for retinal vessel segmentation

Retinal vessel segmentation based on deep learning is an important auxiliary method for assisting clinical doctors in diagnosing retinal diseases. However, existing methods often produce mis-segmentation when dealing with low contrast images and thin blood vessels, which affects the continuity and integrity of the vessel skeleton. In addition, existing deep learning methods tend to lose a lot of detailed information during training, which affects the accuracy of segmentation. To address these issues, we propose a novel dual-decoder based Cross-patch Feature Interactive Net with Edge Refinement (CFI-Net) for end-to-end retinal vessel segmentation. In the encoder part, a joint refinement down-sampling method (JRDM) is proposed to compress feature information in the process of reducing image size, so as to reduce the loss of thin vessels and vessel edge information during the encoding process. In the decoder part, we adopt a dual-path model based on edge detection, and propose a Cross-patch Interactive Attention Mechanism (CIAM) in the main path to enhancing multi-scale spatial channel features and transferring cross-spatial information. Consequently, it improve the network’s ability to segment complete and continuous vessel skeletons, reducing vessel segmentation fractures. Finally, the Adaptive Spatial Context Guide Method (ASCGM) is proposed to fuse the prediction results of the two decoder paths, which enhances segmentation details while removing part of the background noise. We evaluated our model on two retinal image datasets and one coronary angiography dataset, achieving outstanding performance in segmentation comprehensive assessment metrics such as AUC and CAL. The experimental results showed that the proposed CFI-Net has superior segmentation performance compared with other existing methods, especially for thin vessels and vessel edges. The code is available at https://github.com/kita0420/CFI-Net.

S2DA-Net: Spatial and spectral-learning double-branch aggregation network for liver tumor segmentation in CT images

Accurate liver tumor segmentation is crucial for aiding radiologists in hepatocellular carcinoma evaluation and surgical planning. While convolutional neural networks (CNNs) have been successful in medical image segmentation, they face challenges in capturing long-term dependencies among pixels. On the other hand, Transformer-based models demand a high number of parameters and involve significant computational costs. To address these issues, we propose the Spatial and Spectral-learning Double-branched Aggregation Network (S2DA-Net) for liver tumor segmentation. S2DA-Net consists of a double-branched encoder and a decoder with a Group Multi-Head Cross-Attention Aggregation (GMCA) module, Two branches in the encoder consist of a Fourier Spectral-learning Multi-scale Fusion (FSMF) branch and a Multi-axis Aggregation Hadamard Attention (MAHA) branch. The FSMF branch employs a Fourier-based network to learn amplitude and phase information, capturing richer features and detailed information without introducing an excessive number of parameters. The FSMF branch utilizes a Fourier-based network to capture amplitude and phase information, enriching features without introducing excessive parameters. The MAHA branch incorporates spatial information, enhancing discriminative features while minimizing computational costs. In the decoding path, a GMCA module extracts local information and establishes long-term dependencies, improving localization capabilities by amalgamating features from diverse branches. Experimental results on the public LiTS2017 liver tumor datasets show that the proposed segmentation model achieves significant improvements compared to the state-of-the-art methods, obtaining dice per case (DPC) 69.4 % and global dice (DG) 80.0 % for liver tumor segmentation on the LiTS2017 dataset. Meanwhile, the pre-trained model based on the LiTS2017 datasets obtain, DPC 73.4 % and an DG 82.2 % on the 3DIRCADb dataset.

Multi-scale adversarial learning with difficult region supervision learning models for primary tumor segmentation

Objective. Recently, deep learning techniques have found extensive application in accurate and automated segmentation of tumor regions. However, owing to the variety of tumor shapes, complex types, and unpredictability of spatial distribution, tumor segmentation still faces major challenges. Taking cues from the deep supervision and adversarial learning, we have devised a cascade-based methodology incorporating multi-scale adversarial learning and difficult-region supervision learning in this study to tackle these challenges. Approach. Overall, the method adheres to a coarse-to-fine strategy, first roughly locating the target region, and then refining the target object with multi-stage cascaded binary segmentation which converts complex multi-class segmentation problems into multiple simpler binary segmentation problems. In addition, a multi-scale adversarial learning difficult supervised UNet (MSALDS-UNet) is proposed as our model for fine-segmentation, which applies multiple discriminators along the decoding path of the segmentation network to implement multi-scale adversarial learning, thereby enhancing the accuracy of network segmentation. Meanwhile, in MSALDS-UNet, we introduce a difficult region supervision loss to effectively utilize structural information for segmenting difficult-to-distinguish areas, such as blurry boundary areas. Main results. A thorough validation of three independent public databases (KiTS21, MSD’s Brain and Pancreas datasets) shows that our model achieves satisfactory results for tumor segmentation in terms of key evaluation metrics including dice similarity coefficient, Jaccard similarity coefficient, and HD95. Significance. This paper introduces a cascade approach that combines multi-scale adversarial learning and difficult supervision to achieve precise tumor segmentation. It confirms that the combination can improve the segmentation performance, especially for small objects (our codes are publicly availabled on https://zhengshenhai.github.io/).

Philosophical Critique and Decoding Path of Rosa's New Alienation

The theory of alienation occupies a very important position in Marx's theoretical system. In the Economic and Philosophic Manuscripts of 1844, Marx criticizes classical political economy in England starting from alienated labor and puts forward four aspects of alienated labor. The Frankfurt School further developed Marx's theory of alienation and conducted comprehensive investigations on technological alienation, cultural alienation, consumer alienation, and other aspects of modern society's culture, politics, and art. In the new turn of late modern society, the fourth generation Frankfurt School scholar Hartmut Rosa, starting from time analysis, analyzes the growth logic and dynamic stability of modern society's overall crisis through the theory of social acceleration critique, and proposes the social pathologies of "acceleration" and the solution of "resonance", attempting to achieve a better life for people in the present era through the adaptive mode of resonance theory. Based on Hartmut Rosa's works and related research, this article explores the theoretical origins of Rosa's theory of social acceleration critique, the historical background and ideological elements, elaborates on Rosa's analysis of the phenomenon of social acceleration, the new forms of alienation under the background of accelerated society, and the resonant solutions to the problem of new alienation. Finally, this article examines the theoretical value and shortcomings of his theory.

Concurrent Learning Approach for Estimation of Pelvic Tilt from Anterior-Posterior Radiograph.

Accurate and reliable estimation of the pelvic tilt is one of the essential pre-planning factors for total hip arthroplasty to prevent common post-operative complications such as implant impingement and dislocation. Inspired by the latest advances in deep learning-based systems, our focus in this paper has been to present an innovative and accurate method for estimating the functional pelvic tilt (PT) from a standing anterior-posterior (AP) radiography image. We introduce an encoder-decoder-style network based on a concurrent learning approach called VGG-UNET (VGG embedded in U-NET), where a deep fully convolutional network known as VGG is embedded at the encoder part of an image segmentation network, i.e., U-NET. In the bottleneck of the VGG-UNET, in addition to the decoder path, we use another path utilizing light-weight convolutional and fully connected layers to combine all extracted feature maps from the final convolution layer of VGG and thus regress PT. In the test phase, we exclude the decoder path and consider only a single target task i.e., PT estimation. The absolute errors obtained using VGG-UNET, VGG, and Mask R-CNN are 3.04 ± 2.49, 3.92 ± 2.92, and 4.97 ± 3.87, respectively. It is observed that the VGG-UNET leads to a more accurate prediction with a lower standard deviation (STD). Our experimental results demonstrate that the proposed multi-task network leads to a significantly improved performance compared to the best-reported results based on cascaded networks.

ETUNet:Exploring efficient transformer enhanced UNet for 3D brain tumor segmentation

Medical image segmentation is a crucial topic in medical image processing. Accurately segmenting brain tumor regions from multimodal MRI scans is essential for clinical diagnosis and survival prediction. However, similar intensity distributions, variable tumor shapes, and fuzzy boundaries pose severe challenges for brain tumor segmentation. Traditional segmentation networks based on UNet struggle to establish explicit long-range dependencies from the feature space due to the limitations of the CNN receptive field. This is particularly crucial for dense prediction tasks such as brain tumor segmentation. Recent works have incorporated the powerful global modeling capability of Transformer into UNet to achieve more precise segmentation results. Nevertheless, these methods encounter some issues: (1) the global information is often modeled by simply stacking Transformer layers for a specific module, resulting in high computational complexity and underutilization of the potential of the UNet architecture; (2) the rich boundary information of tumor subregions in multi-scale features is often overlooked. Motivated by these challenges, we propose an advanced fusion of Transformer with UNet by reexamining the core three parts (encoder, bottleneck, and skip connections). Firstly, we introduce a CNN-Transformer module in the encoder to replace the traditional CNN module, enabling the capture of deep spatial dependencies from input images. To address high-level semantic information, we incorporate a computationally efficient spatial-channel attention layer in the bottleneck for global interaction, highlighting important semantic features from the encoder path output. For irregular lesions, we fuse the multi-scale features from the encoder output and the decoder features in the skip connections by calculating cross-attention. This adaptive querying of valuable information from multi-scale features enhances the boundary localization ability of the decoder path and suppresses redundant features with low correlation. Compared to existing methods, our model further enhances the learning capacity of the overall UNet architecture while maintaining low computational complexity. Experimental results on the BraTS2018 and BraTS2020 datasets for brain tumor segmentation tasks demonstrate that our model achieves comparable or superior results compared to recent CNN or Transformer-based models. The average DSC and HD95 on the two datasets are 0.854, 6.688, and 0.862, 5.455 respectively. At the same time, our model achieves optimal segmentation of Enhancing tumors, showcasing the effectiveness of our method. Our code will be made publicly available at https://github.com/wzhangck/ETUnet.

EDDS-Unet: An Encoder-Decoder Double Skip Connection Scheme for Skin Lesion Segmentation

The paper presents an approach for effective skin lesion segmentation from dermatoscopic images. Aiming at transferring the weights trained from a network originally designed for image classification task, this study proposes to utilize the first layers of EfficientNet as the encoding layers of a U-Net based architecture. Besides, we introduce an encoder-decoder double skip connection scheme, a new skip connection architecture for extracting useful spatial details of skin lesions from the encoding layers. By the double skip scheme, the approach not only fuses information from the encounter layer in the encoder path to the corresponding layer in the decoder path, but also takes into account information of the proceeding encoding layer. In addition, we propose a new decoder network using the Residual blocks and Convolutional Block Attention Module (CBAM) blocks to handle the gradient vanishing problem as well as penalize the weight of each layer. The proposed Encoder-Decoder Double Skip with the Unet architecture, namely EDDS-Unet, has shown promising performance when evaluated on the official ISIC 2017 challenge and the PH2 databases. The proposed method achieves high evaluation scores with the Dice Similarity Coefficients of 0.907 for the ISIC 2017 and 0.950 for the PH2 databases without pre-or post-processing steps.

Mid-DeepLabv3+: A Novel Approach for Image Semantic Segmentation Applied to African Food Dietary Assessments.

Recent decades have witnessed the development of vision-based dietary assessment (VBDA) systems. These systems generally consist of three main stages: food image analysis, portion estimation, and nutrient derivation. The effectiveness of the initial step is highly dependent on the use of accurate segmentation and image recognition models and the availability of high-quality training datasets. Food image segmentation still faces various challenges, and most existing research focuses mainly on Asian and Western food images. For this reason, this study is based on food images from sub-Saharan Africa, which pose their own problems, such as inter-class similarity and dishes with mixed-class food. This work focuses on the first stage of VBDAs, where we introduce two notable contributions. Firstly, we propose mid-DeepLabv3+, an enhanced food image segmentation model based on DeepLabv3+ with a ResNet50 backbone. Our approach involves adding a middle layer in the decoder path and SimAM after each extracted backbone feature layer. Secondly, we present CamerFood10, the first food image dataset specifically designed for sub-Saharan African food segmentation. It includes 10 classes of the most consumed food items in Cameroon. On our dataset, mid-DeepLabv3+ outperforms benchmark convolutional neural network models for semantic image segmentation, with an mIoU (mean Intersection over Union) of 65.20%, representing a +10.74% improvement over DeepLabv3+ with the same backbone.

Enhancing COVID-19 CT Image Segmentation: A Comparative Study of Attention and Recurrence in UNet Models.

Imaging plays a key role in the clinical management of Coronavirus disease 2019 (COVID-19) as the imaging findings reflect the pathological process in the lungs. The visual analysis of High-Resolution Computed Tomography of the chest allows for the differentiation of parenchymal abnormalities of COVID-19, which are crucial to be detected and quantified in order to obtain an accurate disease stratification and prognosis. However, visual assessment and quantification represent a time-consuming task for radiologists. In this regard, tools for semi-automatic segmentation, such as those based on Convolutional Neural Networks, can facilitate the detection of pathological lesions by delineating their contour. In this work, we compared four state-of-the-art Convolutional Neural Networks based on the encoder-decoder paradigm for the binary segmentation of COVID-19 infections after training and testing them on 90 HRCT volumetric scans of patients diagnosed with COVID-19 collected from the database of the Pisa University Hospital. More precisely, we started from a basic model, the well-known UNet, then we added an attention mechanism to obtain an Attention-UNet, and finally we employed a recurrence paradigm to create a Recurrent-Residual UNet (R2-UNet). In the latter case, we also added attention gates to the decoding path of an R2-UNet, thus designing an R2-Attention UNet so as to make the feature representation and accumulation more effective. We compared them to gain understanding of both the cognitive mechanism that can lead a neural model to the best performance for this task and the good compromise between the amount of data, time, and computational resources required. We set up a five-fold cross-validation and assessed the strengths and limitations of these models by evaluating the performances in terms of Dice score, Precision, and Recall defined both on 2D images and on the entire 3D volume. From the results of the analysis, it can be concluded that Attention-UNet outperforms the other models by achieving the best performance of 81.93%, in terms of 2D Dice score, on the test set. Additionally, we conducted statistical analysis to assess the performance differences among the models. Our findings suggest that integrating the recurrence mechanism within the UNet architecture leads to a decline in the model's effectiveness for our particular application.

SAtUNet: Series atrous convolution enhanced U‐Net for lung nodule segmentation

AbstractPrecise and unambiguous segmentation of pulmonary nodules from the CT images is imperative for a CAD framework implementation delineated for the prognosis of lung cancer. Lung nodule segmentation is an appealing research discipline for accurate dismemberment of lung cancer but the irregularity in shades, contours, and compositions, and the affinity between the tumors and the neighboring regions makes it an arduous task. This paper proffers a series atrous convolution enhanced U‐Net which uses a series of concatenated dilated convolution blocks after every stage in the encoder and decoder path. Our approach helps in obtaining the quintessential components from the feature maps, in addition to the absolute convergence of the model. It is largely assessed on the publicly accessible LIDC‐IDRI dataset. The average Dice Similarity Coefficient (DSC) obtained is 81.10% with an Intersection over Union (IoU/ Jaccard Index) of 72.24%. Exploratory outcomes prove that our architecture achieves ameliorate performance.

Two-stage contextual transformer-based convolutional neural network for airway extraction from CT images

Accurate airway segmentation from computed tomography (CT) images is critical for planning navigation bronchoscopy and realizing a quantitative assessment of airway-related chronic obstructive pulmonary disease (COPD). Existing methods face difficulty in airway segmentation, particularly for the small branches of the airway. These difficulties arise due to the constraints of limited labeling and failure to meet clinical use requirements in COPD. We propose a two-stage framework with a novel 3D contextual transformer for segmenting the overall airway and small airway branches using CT images. The method consists of two training stages sharing the same modified 3D U-Net network. The novel 3D contextual transformer block is integrated into both the encoder and decoder path of the network to effectively capture contextual and long-range information. In the first training stage, the proposed network segments the overall airway with the overall airway mask. To improve the performance of the segmentation result, we generate the intrapulmonary airway branch label, and train the network to focus on producing small airway branches in the second training stage. Extensive experiments were performed on in-house and multiple public datasets. Quantitative and qualitative analyses demonstrate that our proposed method extracts significantly more branches and longer lengths of the airway tree while accomplishing state-of-the-art airway segmentation performance. The code is available at https://github.com/zhaozsq/airway_segmentation.

A Multi-Scale Context Aware Attention Model for Medical Image Segmentation.

Medical image segmentation is critical for efficient diagnosis of diseases and treatment planning. In recent years, convolutional neural networks (CNN)-based methods, particularly U-Net and its variants, have achieved remarkable results on medical image segmentation tasks. However, they do not always work consistently on images with complex structures and large variations in regions of interest (ROI). This could be due to the fixed geometric structure of the receptive fields used for feature extraction and repetitive down-sampling operations that lead to information loss. To overcome these problems, the standard U-Net architecture is modified in this work by replacing the convolution block with a dilated convolution block to extract multi-scale context features with varying sizes of receptive fields, and adding a dilated inception block between the encoder and decoder paths to alleviate the problem of information recession and the semantic gap between features. Furthermore, the input of each dilated convolution block is added to the output through a squeeze and excitation unit, which alleviates the vanishing gradient problem and improves overall feature representation by re-weighting the channel-wise feature responses. The original inception block is modified by reducing the size of the spatial filter and introducing dilated convolution to obtain a larger receptive field. The proposed network was validated on three challenging medical image segmentation tasks with varying size ROIs: lung segmentation on chest X-ray (CXR) images, skin lesion segmentation on dermoscopy images and nucleus segmentation on microscopy cell images. Improved performance compared to state-of-the-art techniques demonstrates the effectiveness and generalisability of the proposed Dilated Convolution and Inception blocks-based U-Net (DCI-UNet).

An Improved Multi-Scale Feature Fusion for Skin Lesion Segmentation

Accurate segmentation of skin lesions is still a challenging task for automatic diagnostic systems because of the significant shape variations and blurred boundaries of the lesions. This paper proposes a multi-scale convolutional neural network, REDAUNet, based on UNet3+ to enhance network performance for practical applications in skin segmentation. First, the network employs a new encoder module composed of four feature extraction layers through two cross-residual (CR) units. This configuration allows the module to extract deep semantic information while avoiding gradient vanishing problems. Subsequently, a lightweight and efficient channel attention (ECA) module is introduced during the encoder’s feature extraction stage. The attention module assigns suitable weights to channels through attention learning and effectively captures inter-channel interaction information. Finally, the densely connected atrous spatial pyramid pooling module (DenseASPP) module is inserted between the encoder and decoder paths. This module integrates dense connections and ASPP, as well as multi-scale information fusion, to recognize lesions of varying sizes. The experimental studies in this paper were constructed on two public skin lesion datasets, namely, ISIC-2018 and ISIC-2017. The experimental results show that our model is more accurate in segmenting lesions of different shapes and achieves state-of-the-art performance in segmentation. In comparison to UNet3+, the proposed REDAUNet model shows improvements of 2.01%, 4.33%, and 2.68% in Dice, Spec, and mIoU metrics, respectively. These results suggest that REDAUNet is well-suited for skin lesion segmentation and can be effectively employed in computer-aided systems.