IoU Score Research Articles

Near Real-Time Flood Mapping with Weakly Supervised Machine Learning

Advances in deep learning and computer vision are making significant contributions to flood mapping, particularly when integrated with remotely sensed data. Although existing supervised methods, especially deep convolutional neural networks, have proved to be effective, they require intensive manual labeling of flooded pixels to train a multi-layer deep neural network that learns abstract semantic features of the input data. This research introduces a novel weakly supervised approach for pixel-wise flood mapping by leveraging multi-temporal remote sensing imagery and image processing techniques (e.g., Normalized Difference Water Index and edge detection) to create weakly labeled data. Using these weakly labeled data, a bi-temporal U-Net model is then proposed and trained for flood detection without the need for time-consuming and labor-intensive human annotations. Using floods from Hurricanes Florence and Harvey as case studies, we evaluated the performance of the proposed bi-temporal U-Net model and baseline models, such as decision tree, random forest, gradient boost, and adaptive boosting classifiers. To assess the effectiveness of our approach, we conducted a comprehensive assessment that (1) covered multiple test sites with varying degrees of urbanization, and (2) utilized both bi-temporal (i.e., pre- and post-flood) and uni-temporal (i.e., only post-flood) input. The experimental results showed that the proposed framework of weakly labeled data generation and the bi-temporal U-Net could produce near real-time urban flood maps with consistently high precision, recall, f1 score, IoU score, and overall accuracy compared with baseline machine learning algorithms.

A cGAN-based tumor segmentation method for breast ultrasound images

Objective. This paper proposes a conditional GAN (cGAN)-based method to perform data enhancement of ultrasound images and segmentation of tumors in breast ultrasound images, which improves the reality of the enhenced breast ultrasound image and obtains a more accurate segmentation result. Approach. We use the idea of generative adversarial training to accomplish the following two tasks: (1) in this paper, we use generative adversarial networks to generate a batch of samples with labels from the perspective of label-generated images to expand the dataset from a data enhancement perspective. (2) In this paper, we use adversarial training instead of postprocessing steps such as conditional random fields to enhance higher-level spatial consistency. In addition, this work proposes a new network, EfficientUNet, based on U-Net, which combines ResNet18, an attention mechanism and a deep supervision technique. This segmentation model uses the residual network as an encoder to retain the lost information in the original encoder and can avoid the gradient disappearance problem to improve the feature extraction ability of the model, and it also uses deep supervision techniques to speed up the convergence of the model. The channel-by-channel weighting module of SENet is then used to enable the model to capture the tumor boundary more accurately. Main results. The paper concludes with experiments to verify the validity of these efforts by comparing them with mainstream methods on Dataset B. The Dice score and IoU score reaches 0.8856 and 0.8111, respectively. Significance. This study successfully combines cGAN and optimized EfficientUNet for the segmentation of breast tumor ultrasound images. The conditional generative adversarial network has a good performance in data enhancement, and the optimized EfficientUNet makes the segmentation more accurate.

Instance segmentation method for weed detection using UAV imagery in soybean fields

Weed detection in crops is a new frontier of precision agriculture, which will enable the distinction between desirable and undesirable plants. Accurate and efficient weed detection and discrimination are fundamental for precision weed management. This study presents a novel approach combining the visible colour index with the instance segmentation method based on the encoder and decoder architecture. This approach tackles the challenge of accurately detecting and segmenting weeds amidst the dense distribution of weeds and soybean crops. The colour index was designed to highlight the contrast between plants and soil to mitigate the effects of illumination and background, while the integration of the ResNet101_v and DSASPP in the encoder-decoder architecture served to reinforce the extraction of abundant multi-scale semantic information to elevate the accuracy of weed patch boundary segmentation. Experimental results on real-world field imagery indicated that the proposed approach exhibited superior performance with an accuracy of 0.905 and an IoU score of 0.959 for weeds segmentation, an aAcc of 0.978, mIoU of 0.939 and a mAcc of 0.972 for overall performance. In comparison with Deeplabv3plus, Deeplabv3, FCN, U-net, FastFCN, Swin Transformer, and Vision Transformer, the proposed pipeline achieved comparable performance in terms of segmentation precision and execution time for differentiating between weeds and soybeans. This work represents a noteworthy contribution to the field of precision agriculture, offering a promising approach to the challenging task of weed detection and precision weed control.

Assessing Cardiac Functions of Zebrafish from Echocardiography Using Deep Learning

Zebrafish is a well-established model organism for cardiovascular disease studies in which one of the most popular tasks is to assess cardiac functions from the heart beating echo-videos. However, current techniques are often time-consuming and error-prone, making them unsuitable for large-scale analysis. To address this problem, we designed a method to automatically evaluate the ejection fraction of zebrafish from heart echo-videos using a deep-learning model architecture. Our model achieved a validation Dice coefficient of 0.967 and an IoU score of 0.937 which attest to its high accuracy. Our test findings revealed an error rate ranging from 0.11% to 37.05%, with an average error rate of 9.83%. This method is widely applicable in any laboratory setting and can be combined with binary recordings to optimize the efficacy and consistency of large-scale video analysis. By facilitating the precise quantification and monitoring of cardiac function in zebrafish, our approach outperforms traditional methods, substantially reducing the time and effort required for data analysis. The advantages of our method make it a promising tool for cardiovascular research using zebrafish.

#4694 AUTOMATIC SEGMENTATION OF GLOMERULAR SUBSTRUCTURES BY DEEP LEARNING

Abstract Background and Aims Electron microscopy (EM) complements light microscopy (LM) evaluation of the kidney biopsy. Foot process effacement, as assessed by EM, helps in diagnosing podocytopathies. However, human interpretation of EM images is time-intensive and often subjective. In this pilot, we investigate how deep learning techniques can help in adequate segmentation of the glomerular basement membranes (GBM) and podocytes in EM images. The ultimate goal would be to design an automatic tool for reliable and fast assessment of foot process effacement. Method Podocytes and glomerular basement membranes (GBM) of 10 patients, 5 with podocyte disease and 5 without glomerular changes (where biopsy showed tubulointerstitial pathology on LM) were annotated with a combination of manual annotation and thresholding for providing ground truth. After data preprocessing including splitting, flipping, rotating and tiling, a modified U-net architecture was applied (‘baseline model’). This baseline model was compared to a combination of U-net and a self-supervised contrastive learning approach with pretraining on 100 additional images (SimCLR framework, ‘fine-tuned model’). Segmentation performance was measured by IoU score. Results Segmentation of the glomerular basement membrane was best achieved by the baseline model and resulted in an IoU score of 0.711±0.089 compared to a IoU score of 0.675 ±0.093 in the fine-tuned model Segmentation of the podocytes was most successful in the fine-tuned model, with a IoU score of 0.609±0.118 compared to a IoU score of 0.591 ±0.118 in the baseline model Conclusion This study pioneers in segmenting glomerular substructures on EM images by means of a modified U-net architecture. The next step is training and validation in larger datasets. Data annotation remains a challenge. Inclusion of more images is expected to greatly improve the performance of the model. Guideline 1500-3600 characters

Blood flow characterization in nailfold capillary using optical flow-assisted two-stream network and spatial-temporal image

The blood flow velocity in the nailfold capillary is an important indicator of the status of microcirculation. The conventional manual processing method is both laborious and prone to human artifacts. A feasible way to solve this problem is to use machine learning to assist in image processing and diagnosis. Inspired by the Two-Stream Convolutional Networks, this study proposes an optical flow-assisted two-stream network to segment nailfold blood vessels. Firstly, we use U-Net as the spatial flow network and the dense optical flow as the temporal stream. The results show that the optical flow information can effectively improve the integrity of the segmentation of blood vessels. The overall accuracy is 94.01 %, the Dice score is 0.8099, the IoU score is 0.6806, and the VOE score is 0.3194. Secondly, The flow velocity of the segmented blood vessel is determined by constructing the spatial-temporal (ST) image. The blood flow velocity evaluated is consistent with the typical blood flow speed reported. This study proposes a novel two-stream network for blood vessel segmentation of nailfold capillary images. Combined with ST image and line detection method, it provides an effective workflow for measuring the blood flow velocity of nailfold capillaries.

WRANet: wavelet integrated residual attention U-Net network for medical image segmentation

Medical image segmentation is crucial for the diagnosis and analysis of disease. Deep convolutional neural network methods have achieved great success in medical image segmentation. However, they are highly susceptible to noise interference during the propagation of the network, where weak noise can dramatically alter the network output. As the network deepens, it can face problems such as gradient explosion and vanishing. To improve the robustness and segmentation performance of the network, we propose a wavelet residual attention network (WRANet) for medical image segmentation. We replace the standard downsampling modules (e.g., maximum pooling and average pooling) in CNNs with discrete wavelet transform, decompose the features into low- and high-frequency components, and remove the high-frequency components to eliminate noise. At the same time, the problem of feature loss can be effectively addressed by introducing an attention mechanism. The combined experimental results show that our method can effectively perform aneurysm segmentation, achieving a Dice score of 78.99%, an IoU score of 68.96%, a precision of 85.21%, and a sensitivity score of 80.98%. In polyp segmentation, a Dice score of 88.89%, an IoU score of 81.74%, a precision rate of 91.32%, and a sensitivity score of 91.07% were achieved. Furthermore, our comparison with state-of-the-art techniques demonstrates the competitiveness of the WRANet network.

TC-SegNet: robust deep learning network for fully automatic two-chamber segmentation of two-dimensional echocardiography.

Heart chamber quantification is an essential clinical task to analyze heart abnormalities by evaluating the heart volume estimated through the endocardial border of the chambers. A precise heart chamber segmentation algorithm using echocardiography is essential for improving the diagnosis of cardiac disease. This paper proposes a robust two chamber segmentation network (TC-SegNet) for echocardiography which follows a U-Net architecture and effectively incorporates the proposed modified skip connection, Atrous Spatial Pyramid Pooling (ASPP) modules and squeeze and excitation modules. The TC-SegNet is evaluated on the open-source fully annotated dataset of cardiac acquisitions for multi-structure ultrasound segmentation (CAMUS). The proposed TC-SegNet obtained an average value of F1-score of 0.91, an average Dice score of 0.9284 and an IoU score of 0.8322 which are higher than the reference models used here for comparison. Further, Pixel error (PE) of 1.5109 which are significantly less than the comparison models. The segmentation results and metrics show that the proposed model outperforms the state-of-the-art segmentation methods.

Rice leaf disease detection based on bidirectional feature attention pyramid network with YOLO v5 model

To ensure higher quality, capacity, and production of rice, it is vital to diagnose rice leaf disease in its early stage in order to decrease the usage of pesticides in agriculture which in turn avoids environmental damage. Hence, this article presents a Multi-scale YOLO v5 detection network to detect and classify the rice crop disease in its early stage. The experiment is initially started by pre-processing the rice leaf images obtained from the RLD dataset, after which data set labels are created, which are then divided into train and test sets. DenseNet-201 is used as the backbone network and depth-aware instance segmentation is used to segment the different regions of rice leaf. Moreover, the proposed Bidirectional Feature Attention Pyramid Network (Bi-FAPN) is used for extracting the features from the segmented image and also enhances the detection of diseases with different scales. Furthermore, the feature maps are identified in the detection head, where the anchor boxes are then applied to the output feature maps to produce the final output vectors by the YOLO v5 network. The subset of channels or filters is pruned from different layers of deep neural network models through the principled pruning approach without affecting the full framework performance. The experiments are conducted with RLD dataset with different existing networks to verify the generalization ability of the proposed model. The effectiveness of the network is evaluated based on various parameters in terms of average precision, accuracy, average recall, IoU, inference time, and F1 score, which are achieved at 82.8, 94.87, 75.81, 0.71, 0.017, and 92.45 respectively.

High-Resolution Characterization of Deformation Induced Martensite in Large Areas of Fatigued Austenitic Stainless Steel Using Deep Learning

This paper aims to demonstrate a novel technique enabling the accurate visualization and fast mapping of deformation-induced α′-martensite produced during cyclic straining of a metastable austenitic stainless steel, refined by reversion annealing to different grain sizes. The technique is based on energy and angular separation of the signal electrons in a scanning electron microscope (SEM). Collection of the inelastic backscattered electrons emitted under high take-off angles from a sample surface results in the acquisition of micrographs with high sensitivity to structural defects, such as dislocations, grain boundaries, and other imperfections. The areas with a high density of lattice imperfections reduce the penetration depth of the primary electrons, and simultaneously affect the signal electrons leaving the specimen. This results in an increase in the inelastic backscattered electrons yielded from the vicinity of α′-martensite, and a bright halo surrounds this phase. The α′-martensite phase can thus be separated from the austenitic matrix in SEM micrographs. In this work, we propose a deep learning method for a precise α′-martensite mapping within a large area. Various deep learning-based methods have been tested, and the best result measured by both Dice loss and IoU scores has been achieved using the U-Net architecture extended by the ResNet encoder.

Deep learning-driven end-to-end segmentation of Martian river valleys

Martian river valleys are inextricably linked to Martian researches, including the development of the Martian climate, geological development and shallow water ice distribution. The segmentation of Martian river valleys provides materials for scientific research. Due to the absence of water, however, the features of Martian rivers are not obvious, resulting in inefficient segmentation. In order to realize high-accuracy segmentation, we propose an end-to-end segmentation method of Martian river valleys based on deep learning. We put forward the MDR (multi-scale double residual) convolution module and the TA (triple attention) module to improve Unet, and thus, MDR-Unet-TA. In this network, we replace two 3 × 3 convolution layers in Unet with an MDR convolution module, which uses multi-scale convolution to extract features of multiple sizes, and uses the residual connection to avoid gradient disappearance. In addition, we introduce a TA module into the skip connection, which reduces the feature map difference between the encoder and the decoder, and obtains detailed information during decoding. Experimental results demonstrate that compared with current semantic segmentation networks, MDR-Unet-TA obtains higher accuracy, F1 and IOU scores of 98.78%, 95.77% and 95.12% on the simple test set and 97.50%, 95.12% and 93.50% on the complex test set, respectively. Compared with MultiRes Unet, the improvement by MDR-Unet-Tri1 in accuracy, F1, and IOU is 0.92%, 2.40% and 2.70% respectively on the simple test set, and 1.98%, 1.84% and 3.58% respectively on the complex test set. MDR-Unet-TA improves the segmentation performance significantly compared with the original network, and realizes the end-to-end segmentation of Martian river valleys.

Architecture of Deep Convolutional Encoder-Decoder Networks for Building Footprint Semantic Segmentation

Building extraction from high-resolution aerial images is critical in geospatial applications such as telecommunications, dynamic urban monitoring, updating geographic databases, urban planning, disaster monitoring, and navigation. Automatic building extraction is a massive task because buildings in various places have varied spectral and geometric qualities. As a result, traditional image processing approaches are insufficient for autonomous building extraction from high-resolution aerial imaging applications. Automatic object extraction from high-resolution images has been achieved using semantic segmentation and deep learning models, which have become increasingly important in recent years. In this study, the U-Net model was used for building extraction, initially designed for biomedical image analysis. The encoder part of the U-Net model has been improved with ResNet50, VGG19, VGG16, DenseNet169, and Xception. However, three other models have been implemented to test the performance of the model studied: PSPNet, FPN, and LinkNet. The performance analysis through the intersection of union method has shown that U-Net with the VGG16 encoder presents the best results compared to the other models with a high IoU score of 83.06%. This research aims to examine the effectiveness of these four approaches for extracting buildings from high-resolution aerial data.

A Foreground Prototype-Based One-Shot Segmentation of Brain Tumors.

The potential for enhancing brain tumor segmentation with few-shot learning is enormous. While several deep learning networks (DNNs) show promising segmentation results, they all take a substantial amount of training data in order to yield appropriate results. Moreover, a prominent problem for most of these models is to perform well in unseen classes. To overcome these challenges, we propose a one-shot learning model to segment brain tumors on brain magnetic resonance images (MRI) based on a single prototype similarity score. With the use of recently developed few-shot learning techniques, where training and testing are carried out utilizing support and query sets of images, we attempt to acquire a definitive tumor region by focusing on slices containing foreground classes. It is unlike other recent DNNs that employed the entire set of images. The training of this model is carried out in an iterative manner where in each iteration, random slices containing foreground classes of randomly sampled data are selected as the query set, along with a different random slice from the same sample as the support set. In order to differentiate query images from class prototypes, we used a metric learning-based approach based on non-parametric thresholds. We employed the multimodal Brain Tumor Image Segmentation (BraTS) 2021 dataset with 60 training images and 350 testing images. The effectiveness of the model is evaluated using the mean dice score and mean IoU score. The experimental results provided a dice score of 83.42 which was greater than other works in the literature. Additionally, the proposed one-shot segmentation model outperforms the conventional methods in terms of computational time, memory usage, and the number of data.

A modified U-net-based architecture for segmentation of satellite images on a novel dataset

The remote-sensing-based satellite images have been providing a wealth of information to the scientists for study of environmental changes caused by climate changes or human activities such as destructive cyclones and earthquakes etc. This paper proposes a deep learning-based segmentation model for agriculture images captured from satellites and a novel agriculture-based satellite dataset. The segmentation has been performed on the satellite images into five categories of cultivated land, uncultivated land, residences, water, and forest. The dataset has been created using Sentinel-2 satellite data over the Panipat district in Haryana, India having diversity in crops and land usage. The dataset consists of 16,720 images and their corresponding masks over the years ranging from 2018 to 2020. The proposed model consists of a six-phase encoder-decoder network with a total of 33 convolution layers. The proposed segmentation model has been evaluated on proposed dataset and obtained an efficient metric of 72% IoU score which is better than state-of-the-art models such as U-Net, Link-Net, FPN and DeeplabV3+ score 51%, 46%, 49%, 67% IoU respectively.

Inferior Alveolar Canal Automatic Detection with Deep Learning CNNs on CBCTs: Development of a Novel Model and Release of Open-Source Dataset and Algorithm

Introduction: The need of accurate three-dimensional data of anatomical structures is increasing in the surgical field. The development of convolutional neural networks (CNNs) has been helping to fill this gap by trying to provide efficient tools to clinicians. Nonetheless, the lack of a fully accessible datasets and open-source algorithms is slowing the improvements in this field. In this paper, we focus on the fully automatic segmentation of the Inferior Alveolar Canal (IAC), which is of immense interest in the dental and maxillo-facial surgeries. Conventionally, only a bidimensional annotation of the IAC is used in common clinical practice. A reliable convolutional neural network (CNNs) might be timesaving in daily practice and improve the quality of assistance. Materials and methods: Cone Beam Computed Tomography (CBCT) volumes obtained from a single radiological center using the same machine were gathered and annotated. The course of the IAC was annotated on the CBCT volumes. A secondary dataset with sparse annotations and a primary dataset with both dense and sparse annotations were generated. Three separate experiments were conducted in order to evaluate the CNN. The IoU and Dice scores of every experiment were recorded as the primary endpoint, while the time needed to achieve the annotation was assessed as the secondary end-point. Results: A total of 347 CBCT volumes were collected, then divided into primary and secondary datasets. Among the three experiments, an IoU score of 0.64 and a Dice score of 0.79 were obtained thanks to the pre-training of the CNN on the secondary dataset and the creation of a novel deep label propagation model, followed by proper training on the primary dataset. To the best of our knowledge, these results are the best ever published in the segmentation of the IAC. The datasets is publicly available and algorithm is published as open-source software. On average, the CNN could produce a 3D annotation of the IAC in 6.33 s, compared to 87.3 s needed by the radiology technician to produce a bidimensional annotation. Conclusions: To resume, the following achievements have been reached. A new state of the art in terms of Dice score was achieved, overcoming the threshold commonly considered of 0.75 for the use in clinical practice. The CNN could fully automatically produce accurate three-dimensional segmentation of the IAC in a rapid setting, compared to the bidimensional annotations commonly used in the clinical practice and generated in a time-consuming manner. We introduced our innovative deep label propagation method to optimize the performance of the CNN in the segmentation of the IAC. For the first time in this field, the datasets and the source codes used were publicly released, granting reproducibility of the experiments and helping in the improvement of IAC segmentation.

RFS-Net: Railway Track Fastener Segmentation Network With Shape Guidance

The fastener is one of the main components of a rail track system. In recent years, deep learning methods such as image segmentation have greatly boosted the fastener state detection process. However, there is still a need to improve the segmentation accuracy and speed, especially for the fasteners in complex environments. To handle this problem, a fast and accurate fastener semantic segmentation network named RFS-Net is proposed based on shape guidance, which can offer a better speed/accuracy trade-off performance via a very shallow architecture. Specifically, in the encoder, a two-stream structure (i.e., regular stream and shape stream) that processes the fastener and shape image in parallel is introduced. The shape image is created based on the geometric structure of the fastener, and it is served as input to the shape stream to guide the segmentation of the fastener. The decoder integrates deep features from the two-stream encoder and then recovers the shape information by the shape attention blocks with skipping connections. We provide two versions of RFS-Net: RFS-Net_S (1.0M, 1014FPS) and RFS-Net_L (12.01M, 453FPS) on the NVIDIA RTX 3060. Experimental results demonstrate the effectiveness of our method by achieving a promising trade-off between accuracy and inference speed. In particular, our method is faster and more accurate on a challenging dataset, from fast modes: 1014 FPS for RFS-Net_S versus 724 FPS for Segmenter, to high-quality segmentation: better performance than STDC with nearly one percent (92.36% versus 91.48% Mean IoU score).

BSNet: Boundary-semantic-fusion network for farmland parcel mapping in high-resolution satellite images

Mapping farmland parcels using satellite images is essential for agricultural remote sensing applications. Loss of spatial details and positioning of parcel boundaries are the main challenges in available deep convolution network (DCN) models. This study developed a boundary-semantic-fusion DCN (BSNet) model for delineating farmland parcels from high-resolution satellite images. Central to this method is the combination between shallow-level boundary features with accurate spatial positioning and deep-level semantic features for category identification. First, a general deep convolution framework consisting of boundary, semantic and fusion blocks was implemented for farmland parcel mapping. Second, a particular serial structure with a detaching operation for linking the boundary and semantic blocks was explored to maintain the spatial details and fine-scale boundaries in feature learning. Third, an encoder-decoder fusion block was developed to integrate the boundary and semantic features to produce the final parcel maps. We validated the proposed model with different high-resolution satellite images in two study areas. The experimental results, with improvements greater than 4% in the F1 score and 6% in the IoU score relative to other comparative methods, illustrate the effectiveness of the proposed model for fine-scale farmland parcel mapping.

A Weakly Supervised Semi-Automatic Image Labeling Approach for Deformable Linear Objects

The presence of Deformable Linear Objects (DLOs) such as wires, cables or ropes in our everyday life is massive. However, the applicability of robotic solutions to DLOs is still marginal due to the many challenges involved in their perception. In this letter, a methodology to generate datasets from a mixture of synthetic and real samples for the training of DLOs segmentation approaches is thus presented. The method is composed of two steps. First, key-points along a real-world DLO are labeled by employing a VR tracker operated by a user. Second, synthetic and real-world datasets are mixed for the training of semantic and instance segmentation deep learning algorithms to study the benefit of real-world data in DLOs segmentation. To validate this method a user study and a parameter study are conducted. The results show that the VR tracker labeling is usable as other labeling techniques but reduces the number of clicks. Moreover, mixing real-world and synthetic DLOs data can improve the IoU score of a semantic segmentation algorithm by circa 5%. Therefore, this work demonstrates that labeling real-world data via a VR tracker can be done quickly and, if the real-world data are mixed with synthetic data, the performances of segmentation algorithms for DLOs can be improved.

DPC-MSGATNet: dual-path chain multi-scale gated axial-transformer network for four-chamber view segmentation in fetal echocardiography

Echocardiography is essential in evaluating fetal cardiac anatomical structures and functions when clinicians conduct early treatment and screening for congenital heart defects, a common and intricate fetal malformation. Nevertheless, the prenatal detection rate of fetal CHD remains low since the peculiarities of fetal cardiac structures and the variousness of fetal CHD. Precisely segmenting four cardiac chambers can assist clinicians in analyzing cardiac morphology and further facilitate CHD diagnosis. Hence, we design a dual-path chain multi-scale gated axial-transformer network (DPC-MSGATNet) that simultaneously models global dependencies and local visual cues for fetal ultrasound (US) four-chamber (FC) views and further accurately segments four chambers. Our DPC-MSGATNet includes a global and a local branch that simultaneously operates on an entire FC view and image patches to learn multi-scale representations. We design a plug-and-play module, Interactive dual-path chain gated axial-transformer (IDPCGAT), to enhance the interactions between global and local branches. In IDPCGAT, the multi-scale representations from the two branches can complement each other, capturing the same region’s salient features and suppressing feature responses to maintain only the activations associated with specific targets. Extensive experiments demonstrate that the DPC-MSGATNet exceeds seven state-of-the-art convolution- and transformer-based methods by a large margin in terms of F1 and IoU scores on our fetal FC view dataset, achieving a F1 score of 96.87% and an IoU score of 93.99%. The codes and datasets can be available at https://github.comQiaoSiBo/DPC-MSGATNet.

Lung and Infection CT-Scan-Based Segmentation with 3D UNet Architecture and Its Modification.

COVID-19 is the disease that has spread over the world since December 2019. This disease has a negative impact on individuals, governments, and even the global economy, which has caused the WHO to declare COVID-19 as a PHEIC (Public Health Emergency of International Concern). Until now, there has been no medicine that can completely cure COVID-19. Therefore, to prevent the spread and reduce the negative impact of COVID-19, an accurate and fast test is needed. The use of chest radiography imaging technology, such as CXR and CT-scan, plays a significant role in the diagnosis of COVID-19. In this study, CT-scan segmentation will be carried out using the 3D version of the most recommended segmentation algorithm for bio-medical images, namely 3D UNet, and three other architectures from the 3D UNet modifications, namely 3D ResUNet, 3D VGGUNet, and 3D DenseUNet. These four architectures will be used in two cases of segmentation: binary-class segmentation, where each architecture will segment the lung area from a CT scan; and multi-class segmentation, where each architecture will segment the lung and infection area from a CT scan. Before entering the model, the dataset is preprocessed first by applying a minmax scaler to scale the pixel value to a range of zero to one, and the CLAHE method is also applied to eliminate intensity in homogeneity and noise from the data. Of the four models tested in this study, surprisingly, the original 3D UNet produced the most satisfactory results compared to the other three architectures, although it requires more iterations to obtain the maximum results. For the binary-class segmentation case, 3D UNet produced IoU scores, Dice scores, and accuracy of 94.32%, 97.05%, and 99.37%, respectively. For the case of multi-class segmentation, 3D UNet produced IoU scores, Dice scores, and accuracy of 81.58%, 88.61%, and 98.78%, respectively. The use of 3D segmentation architecture will be very helpful for medical personnel because, apart from helping the process of diagnosing someone with COVID-19, they can also find out the severity of the disease through 3D infection projections.