Fully Convolutional Neural Network Research Articles

Pelvic bone tumor segmentation fusion algorithm based on fully convolutional neural network and conditional random field

Background and objectivePelvic bone tumors represent a harmful orthopedic condition, encompassing both benign and malignant forms. Addressing the issue of limited accuracy in current machine learning algorithms for bone tumor image segmentation, we have developed an enhanced bone tumor image segmentation algorithm. This algorithm is built upon an improved full convolutional neural network, incorporating both the fully convolutional neural network (FCNN-4s) and a conditional random field (CRF) to achieve more precise segmentation. MethodologyThe enhanced fully convolutional neural network (FCNN-4s) was employed to conduct initial segmentation on preprocessed images. Following each convolutional layer, batch normalization layers were introduced to expedite network training convergence and enhance the accuracy of the trained model. Subsequently, a fully connected conditional random field (CRF) was integrated to fine-tune the segmentation results, refining the boundaries of pelvic bone tumors and achieving high-quality segmentation. ResultsThe experimental outcomes demonstrate a significant enhancement in segmentation accuracy and stability when compared to the conventional convolutional neural network bone tumor image segmentation algorithm. The algorithm achieves an average Dice coefficient of 93.31 %, indicating superior performance in real-time operations. ConclusionIn contrast to the conventional convolutional neural network segmentation algorithm, the algorithm presented in this paper boasts a more intricate structure, proficiently addressing issues of over-segmentation and under-segmentation in pelvic bone tumor segmentation. This segmentation model exhibits superior real-time performance, robust stability, and is capable of achieving heightened segmentation accuracy.

MT2DInv-Unet: A 2D magnetotelluric inversion method based on deep-learning technology

Traditional gradient-based inversion methods usually suffer from the problems of falling into local minima and relying heavily on initial guesses. Deep-learning methods have received increasing attention due to their excellent nonlinear fitting ability. However, given the recent application of deep-learning methods in the field of magnetotelluric (MT) inversion, there are currently challenges associated with achieving high inversion resolution and extracting sufficient features. We develop a neural network model (called MT2DInv-Unet) based on the deformable convolution for 2D MT inversion to approximate the nonlinear mapping from the MT response data to the resistivity model. The deformable convolution is achieved by adding an offset to each sample point of the conventional convolution operation, which extracts hidden relationships and allows the flexible adjustment of the size and shape of the feature region. Meanwhile, we design the network structure with multiscale residual blocks, which effectively extract the multiscale features of the MT response data. This design not only enhances the network performance but also alleviates issues such as vanishing gradients and network degradation. The results of synthetic models indicate that our network inversion method has stable convergence, good robustness, and generalization performance, and it performs better than the fully convolutional neural network and U-Net network. Finally, the inversion results of field data show that MT2DInv-Unet can effectively obtain a reliable underground resistivity structure and has a good application prospect in MT inversion.

Recoding double-phase holograms with the full convolutional neural network

Herein, we proposed a method to recode double-phase holograms (DPHs) with a full convolutional neural network (FCN) for alleviating the fringes and spatial shifting noises in that. There are many fringes near the edge of the diffraction field due to the incomplete Fresnel zone plate, and the pixel-by-pixel encoding method of DPH will cause above fringes to appear in the reconstructed image of that as well. Besides, the spatial shifting noises generated during the conversion from complex-amplitude information to pure-phase information will also decrease the reconstruction quality and definition. Depending on the great nonlinear fitting ability of FCN, the proposed method can recode DPHs, alleviating the fringes and removing the spatial shifting noises effectively. Finally, the phase-only holograms with the resolution 2400 × 4096 have been generated in 0.06 s, which have a peak signal to noise ratio (PSNR) of 35.6 dB in the mean reconstruction quality. Optical results showed that compared with the conventional methods, the target images reconstructed by the proposed method have less noise and a higher display quality.

Multi-network approach for image segmentation in non-contrast enhanced cardiac 3D MRI of arrhythmic patients

Left atrial appendage (LAA) is the source of thrombi formation in more than 90% of strokes in patients with nonvalvular atrial fibrillation. Catheter-based LAA occlusion is being increasingly applied as a treatment strategy to prevent stroke. Anatomical complexity of LAA makes percutaneous occlusion commonly performed under transesophageal echocardiography (TEE) and X-ray (XR) guidance especially challenging. Image fusion techniques integrating 3D anatomical models derived from pre-procedural imaging into the live XR fluoroscopy can be applied to guide each step of the LAA closure. Cardiac magnetic resonance (CMR) imaging gains in importance for radiation-free evaluation of cardiac morphology as alternative to gold-standard TEE or computed tomography angiography (CTA). Manual delineation of cardiac structures from non-contrast enhanced CMR is, however, labor-intensive, tedious, and challenging due to the rather low contrast. Additionally, arrhythmia often impairs the image quality in ECG synchronized acquisitions causing blurring and motion artifacts. Thus, for cardiac segmentation in arrhythmic patients, there is a strong need for an automated image segmentation method. Deep learning-based methods have shown great promise in medical image analysis achieving superior performance in various imaging modalities and different clinical applications. Fully-convolutional neural networks (CNNs), especially U-Net, have become the method of choice for cardiac segmentation. In this paper, we propose an approach for automatic segmentation of cardiac structures from non-contrast enhanced CMR images of arrhythmic patients based on CNNs implemented in a multi-stage pipeline. Two-stage implementation allows subdividing the task into localization of the relevant cardiac structures and segmentation of these structures from the cropped sub-regions obtained from previous step leading to efficient and effective way of automated cardiac segmentation.

Evaluation of brain age changes in patients with liver cirrhosis and hepatic encephalopathy with deep learning models based on structural magnetic resonance imaging

Objective: To investigate the brain aging in patients with cirrhosis and hepatic encephalopathy(HE), constructed a prediction model of brain age based on deep learning and T1 high-resolution MRI, and try to reveal the specific regions where cirrhosis and HE accelerating brain aging. Methods: A cross-sectional study. A brain age prediction model based on the 3D full convolutional neural network was constructed through T1 high-resolution MRI data from 3 609 healthy individuals across eight global public datasets. The mean absolute error (MAE) between actual age and predicted brain age, Pearson correlation coefficient (r) and determination coefficient (R2) were calculated to evaluate the accuracy of the model's predictions. A test set (n=555) from the Human Connectome Project was used to assess the accuracy of the model. A total of 136 patients with cirrhosis were recruited from Tianjin First Central Hospital as the case group (79 patients with cirrhosis without HE and 57 patients with cirrhosis with HE), and 70 healthy individuals were recruited from the society as the healthy control group during the same period. Brain-predicted age difference (Brain-PAD), digital connection-A (NCT-A) and digital-symbol test (DST) scores of all subjects were calculated for all subjects to assess brain aging and cognitive function in the healthy control group, the cirrhosis without HE group, and the cirrhosis with HE group. The network occlusion sensitivity analysis method was employed to assess the importance of each brain region in predicting brain age. Results: As for the prediction model, in the training set, MAE=2.85, r=0.98, R2=0.96. In the test set, MAE=4.45, r=0.96, R2=0.92. In the local data set of the healthy control group, MAE=3.77, r=0.85, R2=0.73. The time of NCT-A in both cirrhosis groups was longer than healthy control group, while the DST scores were lower than healthy control group, and the differences were statistically significant (both P<0.001); the Brain-PAD of healthy control group was (0.8±4.5) years, the Brain-PAD of no-HE group was (6.9±8.1) years, and the HE group was (10.2±7.7) years. The differences between the three groups were statistically significant (P<0.001), and the differences between any two groups were statistically significant (all P<0.05). The importance ratio of visual network in predicting brain age increased in cirrhosis patients, and the HE group was higher than no-HE group. Conclusions: In patients with cirrhosis, the cognitive function is reduced, brain aging is accelerated, and these changes are more obvious in patients with HE. The importance differences of each brain network in predicting brain aging provide a new direction for identifying the specific regions where cirrhosis and HE accelerate brain aging.

Cardiac abnormalities from 12‐Lead ECG signals prediction based on deep convolutional neural network optimized with nomadic people optimization algorithm

SummaryCardiovascular disease (CVD) is a most dangerous disease in the world. Early accurate and automated identification helps the medical professional make a correct diagnosis and administer fast treatment and saving many lives. Several studies have been suggested in this area, but no one yield the expected outcomes owing to data imbalance issue in the medical and healthcare industries. To overcome this problem, a Deep Convolutional Neural Network Optimized with Nomadic People Optimization for Cardiac Abnormalities from 12‐Lead ECG Signals Prediction (CCA‐12L ECG‐DCNN‐NPO) is proposed in this manuscript. At first, the input data is pre‐processed under Morphological filtering and Extended Empirical wavelet transformation (MF‐EEWT) for removing the noise. Then one hot encoding technique is used to improve the predictions and classification accuracy of the method. Afterward, Residual Exemplars Local Binary Pattern (RELBP) based Feature extraction is used to extract the morphological and statistical features. These extracted features are given to DCNN classifier. It contains fully convolutional neural network (FCN) and encoder with decoder framework, which activates pixel‐wise categorization to exactly identify Cardiac abnormalities from 12‐Lead ECG signals. The visual geometry group network (VGGNet) is considered as a backbone of FCN for end‐to‐end training. Generally, DCNN method does not adopt any optimization modes to define the optimum parameters and to assure exact detection. Therefore, Nomadic People Optimization (NPO) is considered to enhance the DCNN weight parameters. The CCA‐12L ECG‐DCNN‐NPO technique is implemented in python and the efficacy is analyzed under performance metrics, such as sensitivity, precision, F‐Score, specificity, accuracy and error rate. From the analysis, the proposed technique attains higher accuracy 27.5%, 10.32%, and 16.65%, higher f‐score 30.93%, 11.14% and 15.3%, lower error rate 36.31%, 15.78%, and 28.08% compared with the existing methods, such as Detecting Cardiac Abnormalities from 12‐lead ECG Signals Under Feature Selection, Feature Extraction, and deep Learning Classification (CCA‐12L ECG‐RFC), Channel self‐attention deep learning framework for multi‐cardiac abnormality diagnosis from varied‐lead ECG signals (CCA‐12L ECG‐CSA‐DNN) and Cardiac disease categorization by electrocardiogram sensing utilizing deep neural network (CCA‐12L ECG‐DNN) respectively.

A brain tumor identification using convolution neural network and fully convolution neural network

Brain tumor identification, along with an investigation, is harmful to the patient. Segmentation, therefore, of paying attention to near-neighborhood growth remains accurate, effective, and healthy. Fully Convolution Neural Network (FCNN) is a reliable picture model to capitulate the hide quality. The form of the multifaceted with the incessant pixels taught with the crest state and the symbolic picture taught. In this research, the making of a totally convoluted method to obtain the participation of a random element and the production of correspondingly large-scale output with a resourceful assumption and information.. The approach has had several difficulties, as measurements are accurate for a variety of images. The improvement in the mortality rate of the programmed order is a critical condition. The scheduling of the mind tumor is an exceedingly troublesome task in the exceptional spatial and basic fluctuation that accompanies the local brain tumor. In this research, a programmed detection of Brain tumors proposed using the characterization of CNN. The most critical method of construction is the completion of the use of small holes. CNN's has less predictability and 97.5 accuracies.

Identification of traditional settlement landscape areas in Hainan based on unmanned aerial remote sensing imagery

Abstract To enhance the accuracy of recognizing traditional settlement landscapes in Hainan, this study introduces a landscape recognition model predicated on a full convolutional neural network (FCN). The research delineates the selection of specific types and sensors for Unmanned Aerial Vehicles (UAVs), employing UAV remote sensing technology to capture image data of Hainan's traditional settlement landscapes. The collected initial image data, laden with interference information, underwent a preprocessing step. On this foundation, the landscape recognition model for Hainan's traditional settlements was developed utilizing the FCN. The empirical results derived from deploying the model reveal the presence of ancient buildings within a 400-meter radius of the principal river. Notably, these structures predominantly span a distance ranging from 100 to 350 meters. The spatial distribution pattern of these ancient edifices notably centers around the Zong ancestral hall. Furthermore, when compared to other benchmark models, the proposed FCN model exhibits superior performance in recognizing forest, grassland, and farmland within the Hainan settlement landscape, achieving average recognition rates of 88.66% and 84.91%, respectively. This investigation underscores the significant potential for applying UAV remote sensing technology in identifying traditional settlement landscapes. It provides pivotal technical support and a reference point for the survey of forest resources and ecological monitoring, thereby enhancing the applicability and dissemination of UAV technology in landscape recognition tasks.

Cascaded regression with dual CNN frame work for time effective detection of gliomas cancers

The determination of brain tumor growth primarily relies on the histopathological examination of biopsy samples. Tumor segmentation in the brain presents a significant challenge in medical image analysis due to its complexity. The ultimate goal is to accurately identify and isolate tumor regions. For the segmentation of brain tumors, a variety of deep-learning techniques have been developed, and they have produced promising results. However, achieving accurate segmentation requires the integration of multiple image modalities with varying contrasts. This makes manual segmentation impractical for larger studies, despite its accuracy. Deep learning's exceptional performance has made it an attractive method for quantitative analysis. The field of medical image analysis presents distinctive challenges that must be overcome to achieve optimal results. The ongoing strategy is obtrusive, tedious and inclined to manual mistakes. These weaknesses show that it is so fundamental to play out a completely computerized technique for the multi-characterization of cerebrum cancers in view of deep learning. Thus, this paper presents an efficient time-optimized and deep-learning model based on cascade regression (DLCR) to segment the tumor grade in the following stages: Data Acquisition in which data were obtained from the well-known brain repository BRATS2017, which included 215 HGG (High-Grade Gliomas) and 80 LGG (Low-Grade Gliomas) glioma cases. Fully Convolutional Neural Network (FCNN) preprocessing was used to remove noise and anomalies from the raw data, and Gaussian Mixture Model feature extraction was used to extract features from the preprocessed image and finally the proposed DLCR model for grade identification. Experimental findings indicate that the suggested system surpasses other pre-existing models in various aspects (accuracy: 0.96, sensitivity:0.97, precision:0.88).

Deep reinforcement learning with light-weight vision model for sequential robotic object sorting

Training an agent to generate cooperative joint learning policies for executing object sorting in cluttered scenes is complex. This paper presents a comparative model-free deep reinforcement learning (DRL) system built from variants of end-to-end lightweight deep neural networks that jointly complement three primitive policies (pushing, grasping, and placing) that operate from visual observations to actions. The end goal of the proposed DRL systems is to perform object sorting in very dynamic and complex environmental scenarios of cluttered and random objects (irregular and regular blocks) existing in different color variations. For this, we investigated DRL methods built from 12 custom instances of Pixel-wise Q-valued Critic Networks (PQCN) from four main backbone networks (DenseNet121, DenseNet169, MobileNetV3, and SqueezeNet) individually combined with custom fully convolutional neural networks (FCN) for learning the image-based observation space through affordance mapping paradigm while considering both dual and single transfer learning schemes. Additionally, we explored three categories of gradient-based optimization methods and considered custom reward functions with varying discount factors. The results show that the PQCN constructed by integrating DenseNet121 or MobileNetV3 with FCN and trained by a dual transfer learning scheme (with a discount factor of γ=0.70) yielded the best performance in most categories of our evaluation in the training phases. Overall, the PQCN-DenseNet121, when trained with a dual transfer learning process, yielded the best generalization (sorting success rate) in most of the evaluation metrics across all object categories in the testing phase. Furthermore, our study provides a new benchmark for validating deep reinforcement learning algorithms while sorting object blocks under varying degrees of complexity. This research finds practical relevance in industrial fields such as manufacturing and packaging, building construction, plant sorting, automobile assembly lines, and waste sorting. Here is the article’s source code: https://github.com/Emmanuel-Okafor/DRL_for_Object_Sorting.git.

Revolutionizing building damage detection: A novel weakly supervised approach using high-resolution remote sensing images

ABSTRACT Rapidly estimating post-disaster building damage via high-resolution remote sensing (HRRS) imagery is essential for initial disaster relief. However, the complex appearance of building damage poses challenges for existing methods. Specifically, relying solely on post-disaster images lacks building boundary guidance, while change detection methods using dual-temporal imageries are prone to introducing false changes. To address these issues, this paper presents a novel weakly supervised approach that leverages pre- and post-disaster HRRS images for building damage detection. The contributions of this paper are twofold. Firstly, a unique framework is proposed to utilize dual-temporal images. Precisely, the proposed method initially extracts fine-grained sub-building-level individuals from pre-disaster images by combining a fully convolutional neural network (FCN)-based method with superpixel segmentation. Then, these details serve as cues to effectively guide the detection of damaged building areas on post-disaster images, thereby enhancing accuracy. Secondly, we propose a weakly supervised method that solely relies on labeling building damage based on image patches but can ultimately yield pixel-level building damage results. Experiments conducted using HRRS images captured during the 2010 Haiti earthquake demonstrate that the proposed method outperforms existing methodologies. This effort of this paper will contribute to the sustainable development of cities and human settlements.

Evaluation of mediastinal lymph node segmentation of heterogeneous CT data with full and weak supervision

Accurate lymph node size estimation is critical for staging cancer patients, initial therapeutic management, and assessing response to therapy. Current standard practice for quantifying lymph node size is based on a variety of criteria that use uni-directional or bi-directional measurements. Segmentation in 3D can provide more accurate evaluations of the lymph node size. Fully convolutional neural networks (FCNs) have achieved state-of-the-art results in segmentation for numerous medical imaging applications, including lymph node segmentation. Adoption of deep learning segmentation models in clinical trials often faces numerous challenges. These include lack of pixel-level ground truth annotations for training, generalizability of the models on unseen test domains due to the heterogeneity of test cases and variation of imaging parameters. In this paper, we studied and evaluated the performance of lymph node segmentation models on a dataset that was completely independent of the one used to create the models. We analyzed the generalizability of the models in the face of a heterogeneous dataset and assessed the potential effects of different disease conditions and imaging parameters. Furthermore, we systematically compared fully-supervised and weakly-supervised methods in this context. We evaluated the proposed methods using an independent dataset comprising 806 mediastinal lymph nodes from 540 unique patients. The results show that performance achieved on the independent test set is comparable to that on the training set. Furthermore, neither the underlying disease nor the heterogeneous imaging parameters impacted the performance of the models. Finally, the results indicate that our weakly-supervised method attains 90%− 91% of the performance achieved by the fully supervised training.

An AI approach to operationalise global daily PlanetScope satellite imagery for river water masking

Monitoring rivers is vital to manage the invaluable ecosystem services they provide, and also to mitigate the risks they pose to property and life through flooding and drought. Due to the vast extent and dynamic nature of river systems, Earth Observation (EO) is one of the best ways to measure river characteristics. As a first step, EO-based river monitoring often requires extraction of accurate pixel-level water masks, but satellite images traditionally used for this purpose suffer from limited spatial and/or temporal resolution. We address this problem by applying a novel Convolutional Neural Network (CNN)-based model to automate water mask extraction from daily 3 m resolution PlanetScope satellite imagery. Notably, this approach overcomes radiometric issues that frequently present limitations when working with CubeSat data. We test our classification model on 36 rivers across 12 global terrestrial biomes (as proxies for the environmental and physical characteristics that lead to the variability in catchments around the globe). Using a relatively shallow CNN classification model, our approach produced a median F1 accuracy score of 0.93, suggesting that a compact and efficient CNN-based model can work as well as, if not better than, the very deep neural networks conventionally used in similar studies, whilst requiring less training data and computational power. We further show that our model, specialised to the task at hand, performs better than a state-of-the-art Fully Convolutional Neural Network (FCN) that struggles with the highly variable image quality from PlanetScope. Although classifying rivers that were narrower than 60 m, anastomosed or highly urbanised was slightly less successful than our other test images, we showed that fine tuning could circumvent these limitations to some degree. Indeed, fine tuning carried out on the Ottawa River, Canada, by including just 5 additional site-specific training images significantly improved classification accuracy (F1 increased from 0.81 to 0.90, p < 0.01). Overall, our results show that CNN-based classification applied to PlanetScope imagery is a viable tool for producing accurate, temporally dynamic river water masks, opening up possibilities for river monitoring investigations where high temporal variability data is essential.

ROOF3D: A REAL AND SYNTHETIC DATA COLLECTION FOR INDIVIDUAL BUILDING ROOF PLANE AND BUILDING SECTIONS DETECTION

Abstract. Deep learning is a powerful tool to extract both individual building and roof plane polygons. But deep learning requires a large amount of labeled data. Hence, publicly available level of detail (LoD)-2 datasets are a natural choice to train fully convolutional neural networks (FCNs) models for both building section and roof plane instance segmentation. Since publicly available datasets are often automatically derived, e.g. based on laser scanning, they lack on annotation accuracy. To complement such a dataset, we introduce manually annotated and synthetically generated data. Manually annotated data is domain-specific and has a high annotation quality but is expensive to obtain. Synthetically generated data has high-quality annotations by definition, but lacks domain-specificity. Moreover, we not only detect individual building section instances, but also roof plane instances. We predict separations not only between individual buildings, but also by a class that describes the line which separates roof planes. The predicted building and roof plane instances are polygonized by a simple tree search algorithm. To achieve more regular polygons, we utilize the Douglas-Peucker polygon simplification algorithm. We describe our dataset in detail to allow comparability between successive methods. To facilitate future works in building and roof plane prediction, our Roof3D dataset is accessible at &lt;code&gt;https://github.com/dlrPHS/GPUB&lt;/code&gt;.

Neural network based generation of a 1-dimensional stochastic field with turbulent velocity statistics

We define and study a fully-convolutional neural network stochastic model, NN-Turb, which generates a 1-dimensional field with some turbulent velocity statistics. In particular, the generated process satisfies the Kolmogorov 2/3 law for the second order structure function. It also presents negative skewness across scales (i.e. Kolmogorov 4/5 law) and exhibits intermittency as characterized by skewness and flatness. Furthermore, our model is never in contact with turbulent data and only needs the desired statistical behavior of the structure functions across scales for training.

Single energy CT-based mass density and relative stopping power estimation for proton therapy using deep learning method

BackgroundThe number of patients undergoing proton therapy has increased in recent years. Current treatment planning systems (TPS) calculate dose maps using three-dimensional (3D) maps of relative stopping power (RSP) and mass density. The patient-specific maps of RSP and mass density were obtained by translating the CT number (HU) acquired using single-energy computed tomography (SECT) with appropriate conversions and coefficients. The proton dose calculation uncertainty of this approach is 2.5%-3.5% plus 1 mm margin. SECT is the major clinical modality for proton therapy treatment planning. It would be intriguing to enhance proton dose calculation accuracy using a deep learning (DL) approach centered on SECT.ObjectivesThe purpose of this work is to develop a deep learning method to generate mass density and relative stopping power (RSP) maps based on clinical single-energy CT (SECT) data for proton dose calculation in proton therapy treatment.MethodsArtificial neural networks (ANN), fully convolutional neural networks (FCNN), and residual neural networks (ResNet) were used to learn the correlation between voxel-specific mass density, RSP, and SECT CT number (HU). A stoichiometric calibration method based on SECT data and an empirical model based on dual-energy CT (DECT) images were chosen as reference models to evaluate the performance of deep learning neural networks. SECT images of a CIRS 062M electron density phantom were used as the training dataset for deep learning models. CIRS anthropomorphic M701 and M702 phantoms were used to test the performance of deep learning models.ResultsFor M701, the mean absolute percentage errors (MAPE) of the mass density map by FCNN are 0.39%, 0.92%, 0.68%, 0.92%, and 1.57% on the brain, spinal cord, soft tissue, bone, and lung, respectively, whereas with the SECT stoichiometric method, they are 0.99%, 2.34%, 1.87%, 2.90%, and 12.96%. For RSP maps, the MAPE of FCNN on M701 are 0.85%, 2.32%, 0.75%, 1.22%, and 1.25%, whereas with the SECT reference model, they are 0.95%, 2.61%, 2.08%, 7.74%, and 8.62%. ConclusionThe results show that deep learning neural networks have the potential to generate accurate voxel-specific material property information, which can be used to improve the accuracy of proton dose calculation.Advances in knowledgeDeep learning-based frameworks are proposed to estimate material mass density and RSP from SECT with improved accuracy compared with conventional methods.

Thunderstorm prediction during pre-tactical air-traffic-flow management using convolutional neural networks

Thunderstorms can be a large source of disruption for European air-traffic management causing a chaotic state of operation within the airspace system. In current practice, air-traffic managers are provided with imprecise forecasts which limit their ability to plan strategically. As a result, weather mitigation is performed using tactical measures with a time horizon of three hours. Increasing the lead time of thunderstorm predictions to the day before operations could help air-traffic managers plan around weather and improve the efficiency of air-traffic-management operations. Emerging techniques based on machine learning have provided promising results, partly attributed to reduced human bias and improved capacity in predicting thunderstorms purely from numerical weather prediction data. In this paper, we expand on our previous work on thunderstorm forecasting, by applying convolutional neural networks (CNNs) to exploit the spatial characteristics embedded in the weather data. The learning task of predicting convection is formulated as a binary-classification problem based on satellite data. The performance of multiple CNN-based architectures, including a fully-convolutional neural network (FCN), a CNN-based encoder–decoder, a U-Net, and a pyramid-scene parsing network (PSPNet) are compared against a multi-layer-perceptron (MLP) network. Our work indicates that CNN-based architectures improve the performance of point-prediction models, with a fully-convolutional neural-network architecture having the best performance. Results show that CNN-based architectures can be used to increase the prediction lead time of thunderstorms. Lastly, a case study illustrating the applications of convection-prediction models in an air-traffic-management setting is presented.

Circular economy oriented future building information processing: PSO for CNN approach

With the acceleration of urbanization, resource depletion is becoming increasingly serious. The innovative strategy of resource utilization is of great significance to the development of new green building. Therefore, this paper combines artificial intelligence technology and deeply excavates the basic information of the old building images collected by the survey. It determines the specific future building construction information from the direction of sustainable development. Based on the traditional convolutional neural network U-Net model, this paper improves its shortcomings, proposes an improved IU-Net model and optimizes the full convolutional neural network structure based on Particle Swarm Optimization (PSO). This paper analyzes and designs the future building information system platform oriented to circular economy, and it compares the improved IU-Net model with other models through experiments to obtain the accuracy and other indicators. The experimental results indicate that the system using the IU-Net model of full convolutional neural network based on PSO has higher prediction accuracy. It has better functionality for the analysis and decision-making of future building.

CT image segmentation of meat sheep Loin based on deep learning.

There are no clear boundaries between internal tissues in sheep Computerized Tomography images, and it is difficult for traditional methods to meet the requirements of image segmentation in application. Deep learning has shown excellent performance in image analysis. In this context, we investigated the Loin CT image segmentation of sheep based on deep learning models. The Fully Convolutional Neural Network (FCN) and 5 different UNet models were applied in image segmentation on the data set of 1471 CT images including the Loin part from 25 Australian White rams and Dolper rams using the method of 5-fold cross validation. After 10 independent runs, different evaluation metrics were applied to assess the performances of the models. All models showed excellent results in terms evaluation metrics. There were slight differences among the results from the six models, and Attention-UNet outperformed others methods with 0.998±0.009 in accuracy, 4.391±0.338 in AVER_HD, 0.90±0.012 in MIOU and 0.95±0.007 in DICE, respectively, while the optimal value of LOSS was 0.029±0.018 from Channel-UNet, and the running time of ResNet34-UNet is the shortest.

Brain tumor image segmentation based on improved FPN

PurposeAutomatic segmentation of brain tumors by deep learning algorithm is one of the research hotspots in the field of medical image segmentation. An improved FPN network for brain tumor segmentation is proposed to improve the segmentation effect of brain tumor.Materials and methodsAiming at the problem that the traditional full convolutional neural network (FCN) has weak processing ability, which leads to the loss of details in tumor segmentation, this paper proposes a brain tumor image segmentation method based on the improved feature pyramid networks (FPN) convolutional neural network. In order to improve the segmentation effect of brain tumors, we improved the model, introduced the FPN structure into the U-Net structure, captured the context multi-scale information by using the different scale information in the U-Net model and the multi receptive field high-level features in the FPN convolutional neural network, and improved the adaptability of the model to different scale features.ResultsPerformance evaluation indicators show that the proposed improved FPN model has 99.1% accuracy, 92% DICE rating and 86% Jaccard index. The performance of the proposed method outperforms other segmentation models in each metric. In addition, the schematic diagram of the segmentation results shows that the segmentation results of our algorithm are closer to the ground truth, showing more brain tumour details, while the segmentation results of other algorithms are smoother.ConclusionsThe experimental results show that this method can effectively segment brain tumor regions and has certain generalization, and the segmentation effect is better than other networks. It has positive significance for clinical diagnosis of brain tumors.