Search

Globally, monitoring productivity, occupational health, and safety of construction workers has long been a significant concern. To address this issue, there is an urgent need for efficient methods to continuously monitor construction sites and recognize workers' actions in a timely manner. Recently, advances in electronic technology and pose estimation algorithms have made it easier to obtain skeleton and joint trajectories of human bodies. Deep learning algorithms have emerged as robust and automated tools for extracting and processing 3D skeleton information on construction sites, proving effective for workforce action assessment. However, most previous studies on action recognition have primarily focused on single-stream data, which limited the network's ability to capture more comprehensive worker action features. Therefore, this research proposes a Spatial-Temporal Multi-Feature Network (STMF-Net) designed to utilize six 3D skeleton-based features to monitor and capture the movements of construction workers, thereby recognizing their actions. The experimental results demonstrate an accuracy of 79.36%. The significance of this work lies in its potential to enhance management models within the construction industry, ultimately improving workers' health and work efficiency.

Objective: In diabetes mellitus patients, hyperuricemia may lead to the development of diabetic complications, including macrovascular and microvascular dysfunction. However, the level of blood uric acid in diabetic patients is obtained by sampling peripheral blood from the patient, which is an invasive procedure and not conducive to routine monitoring. Therefore, we developed deep learning algorithm to detect noninvasively hyperuricemia from retina photographs and metadata of patients with diabetes and evaluated performance in multiethnic populations and different subgroups., Materials and Methods: To achieve the task of non-invasive detection of hyperuricemia in diabetic patients, given that blood uric acid metabolism is directly related to estimated glomerular filtration rate(eGFR), we first performed a regression task for eGFR value before the classification task for hyperuricemia and reintroduced the eGFR regression values into the baseline information. We trained 3 deep learning models: (1) metadata model adjusted for sex, age, body mass index, duration of diabetes, HbA1c, systolic blood pressure, diastolic blood pressure; (2) image model based on fundus photographs; (3)hybrid model combining image and metadata model. Data from the Shanghai General Hospital Diabetes Management Center (ShDMC) were used to develop (6091 participants with diabetes) and internally validated (using 5-fold cross-validation) the models. External testing was performed on an independent dataset (UK Biobank dataset) consisting of 9327 participants with diabetes., Results: For the regression task of eGFR, in ShDMC dataset, the coefficient of determination (R2) was 0.684±0.07 (95 % CI) for image model, 0.501±0.04 for metadata model, and 0.727±0.002 for hybrid model. In external UK Biobank dataset, a coefficient of determination (R2) was 0.647±0.06 for image model, 0.627±0.03 for metadata model, and 0.697±0.07 for hybrid model. Our method was demonstrably superior to previous methods. For the classification of hyperuricemia, in ShDMC validation, the area, under the curve (AUC) was 0.86±0.013for image model, 0.86±0.013 for metadata model, and 0.92±0.026 for hybrid model. Estimates with UK biobank were 0.82±0.017 for image model, 0.79±0.024 for metadata model, and 0.89±0.032 for hybrid model., Conclusion: There is a potential deep learning algorithm using fundus photographs as a noninvasively screening adjunct for hyperuricemia among individuals with diabetes. Meanwhile, combining patient's metadata enables higher screening accuracy. After applying the visualization tool, it found that the deep learning network for the identification of hyperuricemia mainly focuses on the fundus optic disc region., Competing Interests: Declaration of competing interest The all author declare that they have no conflicts of interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024. Published by Elsevier B.V.)

The analysis of membrane vesicles at the nanoscale level is crucial for advancing the understanding of intercellular communication and its implications for health and disease. Despite their significance, the nanoscale analysis of vesicles at the single particle level faces challenges owing to their small size and the complexity of biological fluids. This new vesicle analysis tool leverages the single-molecule sensitivity of super-resolution microscopy (SRM) and the high-throughput analysis capability of deep-learning algorithms. By comparing classical clustering methods (k-means, DBSCAN, and SR-Tesseler) with deep-learning-based approaches (YOLO, DETR, Deformable DETR, and Faster R-CNN) for the analysis of super-resolution fluorescence images of exosomes, we identified the deep-learning algorithm, Deformable DETR, as the most effective. It showed superior accuracy and a reduced processing time for detecting individual vesicles from SRM images. Our findings demonstrate that image-based deep-learning-enhanced methods from SRM images significantly outperform traditional coordinate-based clustering techniques in identifying individual vesicles and resolving the challenges related to misidentification and computational demands. Moreover, the application of the combined Deformable DETR and ConvNeXt-S algorithms to differently labeled exosomes revealed its capability to differentiate between them, indicating its potential to dissect the heterogeneity of vesicle populations. This breakthrough in vesicle analysis suggests a paradigm shift towards the integration of AI into super-resolution imaging, which is promising for unlocking new frontiers in vesicle biology, disease diagnostics, and the development of vesicle-based therapeutics., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

The Brain Tumor Segmentation (BraTS) Challenge has been a main driver of the development of deep learning (DL) algorithms and provides by far the largest publicly available expert-annotated brain tumour dataset but contains solely preoperative examinations. The aim of our study was to facilitate the use of the BraTS dataset for training DL brain tumour segmentation algorithms for a postoperative setting. To this end, we introduced an automatic conversion of the three-label BraTS annotation protocol to a two-label annotation protocol suitable for postoperative brain tumour segmentation. To assess the viability of the label conversion, we trained a DL algorithm using both the three-label and the two-label annotation protocols. We assessed the models pre- and postoperatively and compared the performance with a state-of-the-art DL method. The DL algorithm trained using the BraTS three-label annotation misclassified parts of 10 out of 41 fluid-filled resection cavities in 72 postoperative glioblastoma MRIs, whereas the two-label model showed no such inaccuracies. The tumour segmentation performance of the two-label model both pre- and postoperatively was comparable to that of a state-of-the-art algorithm for tumour volumes larger than 1 cm 3 . Our study enables using the BraTS dataset as a basis for the training of DL algorithms for postoperative tumour segmentation.

Background: Alkaline earth metal ions are important protein binding ligands in human body, and it is of great significance to predict their binding residues., Results: In this paper, Mg 2+ and Ca 2+ ligands are taken as the research objects. Based on the characteristic parameters of protein sequences, amino acids, physicochemical characteristics of amino acids and predicted structural information, deep neural network algorithm is used to predict the binding sites of proteins. By optimizing the hyper-parameters of the deep learning algorithm, the prediction results by the fivefold cross-validation are better than those of the Ionseq method. In addition, to further verify the performance of the proposed model, the undersampling data processing method is adopted, and the prediction results on independent test are better than those obtained by the support vector machine algorithm., Conclusions: An efficient method for predicting Mg 2+ and Ca 2+ ligand binding sites was presented., (© 2021. The Author(s).)

Recent advances in flexible materials and wearable electronics offer a noninvasive, high-fidelity recording of biopotentials for portable healthcare, disease diagnosis, and machine interfaces. Current device-manufacturing methods, however, still heavily rely on the conventional cleanroom microfabrication that requires expensive, time-consuming, and complicated processes. Here, we introduce an additive nanomanufacturing technology that explores a contactless direct printing of aerosol nanomaterials and polymers to fabricate stretchable sensors and multilayered wearable electronics. Computational and experimental studies prove the mechanical flexibility and reliability of soft electronics, considering direct mounting to the deformable human skin with a curvilinear surface. The dry, skin-conformal graphene biosensor, without the use of conductive gels and aggressive tapes, offers an enhanced biopotential recording on the skin and multiple uses (over ten times) with consistent measurement of electromyograms. The combination of soft bioelectronics and deep learning algorithm allows classifying six classes of muscle activities with an accuracy of over 97%, which enables wireless, real-time, continuous control of external machines such as a robotic hand and a robotic arm. Collectively, the comprehensive study of nanomaterials, flexible mechanics, system integration, and machine learning shows the potential of the printed bioelectronics for portable, smart, and persistent human-machine interfaces.

Purpose: Oral cancer is a complex wide spread cancer, which has high severity. Using advanced technology and deep learning algorithm early detection and classification are made possible. Medical imaging technique, computer-aided diagnosis and detection can make potential changes in cancer treatment. In this research work, we have developed a deep learning algorithm for automated, computer-aided oral cancer detecting system by investigating patient hyperspectral images., Methods: To validate the proposed regression-based partitioned deep learning algorithm, we compare the performance with other techniques by its classification accuracy, specificity, and sensitivity. For the accurate medical image classification objective, we demonstrate a new structure of partitioned deep Convolution Neural Network (CNN) with two partitioned layers for labeling and classify by labeling region of interest in multidimensional hyperspectral image., Results: The performance of the partitioned deep CNN was verified by classification accuracy. We have obtained classification accuracy of 91.4% with sensitivity 0.94 and a specificity of 0.91 for 100 image data sets training for task classification of cancerous tumor with benign and for task classification of cancerous tumor with normal tissue accuracy of 94.5% for 500 training patterns was obtained., Conclusions: We compared the obtained results from another traditional medical image classification algorithm. From the obtained result, we identify that the quality of diagnosis is increased by proposed regression-based partitioned CNN learning algorithm for a complex medical image of oral cancer diagnosis.

The two-phase flow in a microchannel consists of liquid–liquid and gas–liquid material components. The automatic recognition of flow patterns using deep learning approaches has been emerging. This study aimed to improve the recognition accuracy of flow patterns in the two-phase flow images. The different convolutional kernels in the GoogLeNet algorithm extracted the image features with different scales. In order to strengthen the important channel and spatial features, this paper proposes the combined five-layer Coord attention and GoogLeNet algorithm to enhance the accuracy of the new algorithm. The optimized algorithm model was derived from image datasets with different liquid–liquid two-phase flows (NaAlg–Oil, GaInSn–Water), and its accuracy was 95.09% in training and 98.12% in testing. This new model was also applied to predict the flow patterns, with a recognition accuracy of more than 97% in both the liquid–liquid and gas–liquid two-phase flows (water–soybean oil, water–lubricating oil, and argon–water). [ABSTRACT FROM AUTHOR]

In this narrative review, we focus on the application of artificial intelligence in the clinical history of patients with glomerular disease, digital pathology in kidney biopsy, renal ultrasonography imaging, and prediction of chronic kidney disease (CKD). With the development of natural language processing, the clinical history of a patient can be used to identify a computable phenotype. In kidney pathology, digital imaging has adopted innovative deep learning algorithms (DLAs) that can improve the predictive capability of the examined lesions. However, at this time, these applications can only be used in research because there is no recognized validation to replace the conventional diagnostic applications. Kidney ultrasonography, used in the clinical examination of patients, provides information about the progression of kidney damage. Machine learning algorithms (MLAs) with promising results for the early detection of CKD have been proposed, but, still, they are not solid enough to be incorporated into the clinical practice. A few tools for glomerulonephritis, based on MLAs, are available in clinical practice. They can be downloaded on computers and cellular phones but can only be applied to uniracial cohorts of patients. To improve their performance, it is necessary to organize large consortia with multiracial cohorts. Finally, in many studies MLA development has been carried out using retrospective cohorts. The performance of the models might differ in retrospective cohorts compared to real-world data. Therefore, the models should be validated in prospective external large cohorts. [ABSTRACT FROM AUTHOR]

This article proposes the deep learning algorithm- Convolutional Neural Networks (CNN) for both Glioma tumor classifications and diagnosis process. This proposed CNN architecture is derived from the conventional CNN architecture to obtain the optimum classification and diagnosis accuracy. This proposed CNN architecture is derived from the conventional system for obtaining the high classification and diagnosis performance. This proposed methodology stated in this paper uses BRATS 2015 open access dataset for obtaining the brain Magnetic Resonance Image (MRI) for tumor region detection. The proposed methodology stated in this paper for tumor diagnosis achieves 97.7% of Jaccard Index (J) and 83.8% of Dice Similarity Index (DSI) and 99.025 of Diagnosis Rate (DR) using CNN algorithm.. [ABSTRACT FROM AUTHOR]

To reduce the pain and the probability of cross-infection caused by the invasive blood glucose testing instruments, the ex vivo glucose measurement is of high significance. The electrical property of blood varies with the density of the glucose, which can be sensed by measuring its reflected coefficient in millimeter-wave. In this article, we built a contactless glucose solution concentration measurement system based on 77-GHz FMCW radar. Several preliminary signal processing algorithms are cascaded with a deep neural network to improve the accuracy of glucose solution concentration measurement. Our experiment shows that the resolution of this ex vivo glucose measurement can achieve up to 0.1 mg/mL. [ABSTRACT FROM AUTHOR]

To solve the mismatch between heating quantity and demand of thermal stations, an optimized control method based on depth deterministic strategy gradient was proposed in this paper. In this paper, long short-time memory deep learning algorithm is used to model the thermal power station, and then the depth deterministic strategy gradient control algorithm is used to solve the water supply flow sequence of the primary side of the thermal power station in combination with the operation mechanism of the central heating system. In this paper, a large number of historical working condition data of a thermal station are used to carry out simulation experiment, and the results show that the method is effective, which can realize the on-demand heating of the thermal station a certain extent and improve the utilization rate of heat. [ABSTRACT FROM AUTHOR]

• Deep Learning architecture for illumination correction. • Works on RGB images of paintings and other reflective surfaces like laptop screens. • A very simple model, which is low on time-space complexity. • Works on both over exposed and underexposed pictures. Image illumination correction has been a long standing topic for research in the Computer Vision problem. However, all previous literature on this topic has either been statistical in nature in the sense that a specified algorithm has been developed for approaching a particular case of illumination normalization, or involves extremely complex deep learning methods for illumination correction of either one of over illuminated or under illuminated images. We present here a very simple deep learning based image illumination correction architecture which works on color images of paintings irrespective of whether they are under or over illuminated. We have tested the results using a synthetic database as well as on real world painting images of diverse nature. [ABSTRACT FROM AUTHOR]

Summary: Non‐technical loss (NTL) is detrimental to the smart grid. Intelligent application of advanced metering infrastructure (AMI) helps to solve NTL detection and classification. By using advanced learning algorithms, data analysis on the massive data generated by AMI is helpful in the detection and classification of electricity theft. Conventional data analysis algorithm, like Support Vector Machine (SVM), Random Forest Algorithm (RFA), and 1D‐ Conventional Neural Network (1D‐CNN), has low detection and classification accuracy of electricity theft. Because these methods failed to predict and classify multidimensional electricity consumption data by various consumers in AMI in the smart grid. In this research work, a multidimensional deep learning algorithm is proposed to learn and classify the non‐periodicity of electricity. This helps to detect electricity theft by a consumer from the periodic load curve. Both weekly load patterns and daily load patterns are processed as 2D electricity data samples. From the proposed multidimensional deep learning model, an average classification accuracy of 97.5% and a precision‐recall of 0.97 were obtained. This validates that the proposed deep learning model outperforms other conventional data analysis classification algorithm. [ABSTRACT FROM AUTHOR]

Detecting the electrocardiogram pattern in Internet of Things–based healthcare system and notifying this to the user is a challenging task. Using advance computing methods for classification of electrocardiogram signal is a notable research topic. In this research work, an intelligent electrocardiogram signal classification, employing deep learning algorithm, developed and tested in Internet of Things–based smart healthcare system was proposed. For classification of acquired electrocardiogram signal, a partitioned deep convolutional neural network was proposed. The electrocardiogram feature continuously in the Internet of Things–based monitoring system was learnt. To make use of learned features in the continuous time series data, it forms a higher order space in the server. We have made quantifiable comparative analysis with other classification algorithm with the same time series data collected from different atrial fibrillation samples in the Internet of Things–based e-health system. Our proposed algorithm learned features were tested in atrial fibrillation classified signal with other conventional classifiers with various performance indices. We obtained an accuracy of 96.3 percent with 93.5-percent sensitivity and 97.5-percent precision. From the obtained result, processing with proposed deep convolutional neural network provides reliable timely assist and accurate classification of electrocardiogram signal in Internet of Things–based smart healthcare system. [ABSTRACT FROM AUTHOR]

• Pixel level segmentation algorithm of irregular aggregate based on deep learning. • Online measurement system of stacking aggregate gradation containing water and mud. The quality of image processing affects the results of aggregate particle size detection. So far, the aggregate image segmentation is still limited to the traditional watershed algorithm (WA) and threshold segmentation algorithm, and their segmentation effect is limited. Therefore, a deep learning algorithm (DLA) to segment stacked coarse aggregate images was presented in this paper. Then it was used to segment the images of aggregates with different aggregate materials, particle sizes and mixing ratios, and the segmentation results were analyzed. The dataset required was made by the labelme, and trained to obtain the segmentation model of the DLA. The test results were compared with that of the traditional WA. For single-stage aggregates with particle sizes of 5–10 mm, 10–20 mm and 20–31.5 mm, the proportion of cumulative gradation calculated by DLA was increased by 5.93%, 28.62% and 23.62%, respectively, for different base materials of coarse aggregate, the proportion of cumulative gradation calculated by DLA was increased by 28.62%, 37.68% and 12.78%, respectively. For aggregate images with different mixing ratios, the segmentation results of DLA were closer to the actually configured gradation. The experimental results showed that the DLA could avoid the common phenomenon of over-segmentation and under-segmentation in the WA, and had general applicability in segmenting coarse aggregates with different particle sizes and aggregate materials. [ABSTRACT FROM AUTHOR]

An increasing focus was placed in the past few decades on accelerating flash calculations and a variety of acceleration strategies have been developed to improve its efficiency without serious compromise in accuracy and reliability. Recently, as machine learning becomes a powerful tool to handle complicated and time-consuming problems, it is increasingly appealing to replace the iterative flash algorithm, due to the strong nonlinearity of flash problem, by a neural network model. In this study, an NVT flash calculation scheme is established with a thermodynamically stable evolution algorithm to generate training and testing data for the proposed deep neural network. With a modified network structure, the deep learning algorithm is optimized by carefully tuning neural network hyperparameters. Numerical tests indicate that the trained model is capable of accurately estimating phase compositions and states for complex reservoir fluids under a wide range of environmental conditions, while the effect of capillary pressure can be captured well. Thermodynamic rules are preserved well through our algorithm, and the trained model can be used for various fluid mixtures, which significantly accelerates flash calculations in unconventional reservoirs. • · A thermodynamically consistent modeling and algorithm to calculate phase equilibrium • Capillary pressure is taken into consideration in the flash calculation. • An optimized deep neural network is designed to accelerate the flash calculation. • · Our deep learning algorithm is capable of capturing the capillary effect and supercritical fluid phenomena. [ABSTRACT FROM AUTHOR]

• This paper proposes a deep learning algorithm for rapid detection of aortic dissection using contrast-enhanced CT images. • We built a segmentation network that is more suitable for segmenting CT images of the aorta. • This research has laid the foundation for the development of an effective aortic dissection detection algorithm. Aortic dissection (AD) is one of the most common aortic diseases, where blood enters the aortic wall through the aortic intimal rift and causes separation of the arterial wall. Any delay or misdiagnosis can have severe consequences for patients with aortic dissection and even lead to higher mortality rates. Therefore, rapid and accurate detection of aortic dissection saves patients valuable time and provides assistance for the selection of clinical treatment options. This paper describes a deep learning algorithm that uses contrast-enhanced CT images for segmentation and automatic detection of aortic dissection. First, we construct a U-Net based semantic segmentation architecture and apply it to contrast-enhanced CT images to segment the aortic true lumen. Then, we use the segmentation results for aortic circularity analysis to obtain slice-level detection results. Finally, we aggregated the slice-level results to present patient-level detection results. We tested our algorithm on 20 contrast-enhanced CT datasets, of which 10 were aortic dissections. In terms of temporal performance, we have achieved millisecond prediction on sliced images. At the same time, we achieved 85.00% accuracy, 90.00% sensitivity and 80.00% specificity in patient-level testing. [ABSTRACT FROM AUTHOR]

The brain tumor can be created by uncontrollable increase of abnormal cells in tissue of brain and it has two kinds of tumors: one is benign and another one is malignant tumor. The benign brain tumor does not affect the adjacent normal and healthy tissue but the malignant tumor can affect the neighboring tissues of brain that can lead to the death of person. An early detection of brain tumor can be required to protect the survival of patients. Usually, the brain tumor is detected using MRI scanning method. However, the radiologists are not providing the effective tumor segmentation in MRI image due to the irregular shape of tumors and position of tumor in the brain. Accurate brain tumor segmentation is needed to locate the tumor and it is used to give the correct treatment for a patient and it provides the key to the doctor who must execute the surgery for patient. In this paper, a novel deep learning algorithm (kernel based CNN) with M-SVM is presented to segment the tumor automatically and efficiently. This presented work contains several steps that are preprocessing, feature extraction, image classification and tumor segmentation of brain. The MRI image is smoothed and enhanced by Laplacian of Gaussian filtering method (LoG) with Contrast Limited Adaptive Histrogram Equalization (CLAHE) and feature can be extracted from it based on tumor shape position, shape and surface features in brain. Consequently, the image classification is done using M-SVM depending on the selected features. From MRI image, the tumor is segmented with help of kernel based CNN method.. Experimental results of proposed method can show that this presented technique can executes brain tumor segmentation accurately reaching almost 84% in evaluation with existing algorithms. [ABSTRACT FROM AUTHOR]