Search

Microseismic source location is crucial for the early warning of rockburst risks. However, the conventional methods face challenges in terms of the microseismic wave velocity and arrival time accuracy. Intelligent techniques, such as the full convolutional neural network (FCNN), can capture spatial information but struggle with complex microseismic sequence. Combining the FCNN with the long short-term memory (LSTM) network enables better time-series signal classification by integrating multi-scale information and is therefore suitable for waveform location. The LSTM-FCNN model does not require extensive data preprocessing and it simplifies the microseismic source location through feature extraction. In this study, we utilized the LSTM-FCNN as a regression learning model to locate the seismic focus. Initially, the method of short-time-average/long-time-average (STA/LTA) arrival time picking was employed to augment spatiotemporal information. Subsequently, oversampling the on-site data was performed to address the issue of data imbalance, and finally, the performance of LSTM-FCNN was tested. Meanwhile, we compared the LSTM-FCNN model with previous deep-learning models. Our results demonstrated remarkable location capabilities with a mean absolute error (MAE) of only 7.16 m. The model can realize swift training and high accuracy, thereby significantly improving risk warning of rockbursts.

Tiger puffer is a commercially important fish cultured in high-density environments, and its accurate detection is indispensable for determining growth conditions and realizing accurate feeding. However, the detection precision and recall of farmed tiger puffer are low due to target blurring and occlusion in real farming environments. The farmed tiger puffer detection model, called knowledge aggregation YOLO (KAYOLO), fuses prior knowledge with improved YOLOv5 and was proposed to solve this problem. To alleviate feature loss caused by target blurring, we drew on the human practice of using prior knowledge for reasoning when recognizing blurred targets and used prior knowledge to strengthen the tiger puffer's features and improve detection precision. To address missed detection caused by mutual occlusion in high-density farming environments, a prediction box aggregation method, aggregating prediction boxes of the same object, was proposed to reduce the influence among different objects to improve detection recall. To validate the effectiveness of the proposed methods, ablation experiments, model performance experiments, and model robustness experiments were designed. The experimental results showed that KAYOLO's detection precision and recall results reached 94.92% and 92.21%, respectively. The two indices were improved by 1.29% and 1.35%, respectively, compared to those of YOLOv5. Compared with the recent state-of-the-art underwater object detection models, such as SWIPENet, RoIMix, FERNet, and SK-YOLOv5, KAYOLO achieved 2.09%, 1.63%, 1.13% and 0.85% higher precision and 1.2%, 0.18%, 1.74% and 0.39% higher recall, respectively. Experiments were conducted on different datasets to verify the model's robustness, and the precision and recall of KAYOLO were improved by approximately 1.3% compared to those of YOLOv5. The study showed that KAYOLO effectively enhanced farmed tiger puffer detection by reducing blurring and occlusion effects. Additionally, the model had a strong generalization ability on different datasets, indicating that the model can be adapted to different environments, and it has strong robustness.

With the vigorous development of intelligence agriculture, the progress of automated large-scale and intensive pig farming has accelerated significantly. As a biological feature, the pig face has important research significance for precise breeding of pigs and traceability of health. In the management of live pigs, many managers adopt traditional methods, including color marking and RFID identification, but there will be problems such as off-label, mixed-label and waste of manpower. This work proposes a non-invasive way to study the identification of multiple individuals in pigs. The model was to first replace the original backbone network of YOLOv4 with MobileNet-v3, a popular lightweight network. Then depth-wise separable convolution was adopted in YOLOv4′s feature extraction network SPP and PANet to further reduce network parameters. Moreover, CBAM attention mechanism formed by the concatenation of CAM and SAM was added to PANet to ensure the network accuracy while reducing the model weight. The introduction of multi-attention mechanism selectively strengthened key areas of pig face and filtered out weak correlation features, so as to improve the overall model effect. Finally, an improved MobileNetv3-YOLOv4-PACNet (M-YOLOv4-C) network model was proposed to identify individual sows. The mAP were 98.15 %, the detection speed FPS were 106.3frames/s, and the model parameter size was only 44.74 MB, which can be well implanted into the small-volume pig house management sensors and applied to the pig management system in a lightweight, fast and accurate manner. This model will provide model support for subsequent pig behavior recognition and posture analysis.

Objective: To develop a deep learning model to predict lymph node (LN) status in clinical stage IA lung adenocarcinoma patients. Methods: This diagnostic study included 1,009 patients with pathologically confirmed clinical stage T1N0M0 lung adenocarcinoma from two independent datasets (699 from Cancer Hospital of Chinese Academy of Medical Sciences and 310 from PLA General Hospital) between January 2005 and December 2019. The Cancer Hospital dataset was randomly split into a training cohort (559 patients) and a validation cohort (140 patients) to train and tune a deep learning model based on a deep residual network (ResNet). The PLA Hospital dataset was used as a testing cohort to evaluate the generalization ability of the model. Thoracic radiologists manually segmented tumors and interpreted high-resolution computed tomography (HRCT) features for the model. The predictive performance was assessed by area under the curves (AUCs), accuracy, precision, recall, and F1 score. Subgroup analysis was performed to evaluate the potential bias of the study population. Results: A total of 1,009 patients were included in this study; 409 (40.5%) were male and 600 (59.5%) were female. The median age was 57.0 years (inter-quartile range, IQR: 50.0–64.0). The deep learning model achieved AUCs of 0.906 (95% CI: 0.873–0.938) and 0.893 (95% CI: 0.857–0.930) for predicting pN0 disease in the testing cohort and a non-pure ground glass nodule (non-pGGN) testing cohort, respectively. No significant difference was detected between the testing cohort and the non-pGGN testing cohort (P = 0.622). The precisions of this model for predicting pN0 disease were 0.979 (95% CI: 0.963–0.995) and 0.983 (95% CI: 0.967–0.998) in the testing cohort and the non-pGGN testing cohort, respectively. The deep learning model achieved AUCs of 0.848 (95% CI: 0.798–0.898) and 0.831 (95% CI: 0.776–0.887) for predicting pN2 disease in the testing cohort and the non-pGGN testing cohort, respectively. No significant difference was detected between the testing cohort and the non-pGGN testing cohort (P = 0.657). The recalls of this model for predicting pN2 disease were 0.903 (95% CI: 0.870–0.936) and 0.931 (95% CI: 0.901–0.961) in the testing cohort and the non-pGGN testing cohort, respectively. Conclusions: The superior performance of the deep learning model will help to target the extension of lymph node dissection and reduce the ineffective lymph node dissection in early-stage lung adenocarcinoma patients.

This study focuses on developing and evaluating predictive models for bus travel times adaptable to any transit network, or to new roadway segments without prior travel time data. Most prior work relies on non-standardized features such as road traffic forecasts or closed-source datasets to test predictions on a single route or network. We leverage standardized and open-source data from GTFS and GTFS-RT feeds to gather four months of realtime bus position data from Seattle and Trondheim's transit networks. We then test and refine strategies for generalizing model predictions across both locations. To achieve this, we first develop a data pipeline to process and clean the raw data, then extract features from the standardized sources. We then evaluate the performance of several deep learning and heuristic models in predicting bus travel times between source and target bus networks. Holdout data is taken from selected routes in the source city to validate the internal generalization of the models. Data from the target city is used to evaluate the external generalization of the models. An ablation study explores the impact of different open data sources on model generalization (GPS, static timetables, OpenStreetMap and other realtime trips). We then extend the analysis to 33 international bus networks, placing the results in broader context and testing fine-tuning strategies for generalization. Results show that deep learning methods generalize well within the source network, with as little as 1% loss in MAPE on holdout routes. With minimal fine-tuning generalization is significantly improved on the target network. Model features built on static schedule data, realtime positions or OpenStreetMap embeddings improved generalization performance (up to 10% reduction in MAPE). This was more pronounced for networks with a greater initial quantity of training data. As a route-planning tool for roadways without prior data, geospatial data mining can provide reasonable bus travel time estimates. For cross-sectional bus network analysis, fine tuning on at least 100 trajectory samples for each target network is required to significantly outperform baseline heuristics. This necessitates a GTFS-RT or other standardized realtime data feed in the target city.

Introduction:: Accurate identification of pulmonary arterial hypertension (PAH) in primary care and rural areas can be a challenging task. However, recent advancements in computer vision offer the potential for automated systems to detect PAH from echocardiography. Objectives:: Our aim was to develop a precise and efficient diagnostic model for PAH tailored to the unique requirements of intelligent diagnosis, especially in challenging locales like high-altitude regions. Methods:: We proposed the Chamber Attention Network (CAN) for PAH identification from echocardiographic images, trained on a dataset comprising 13,912 individual subjects. A convolutional neural network (CNN) for view classification was used to select the clinically relevant apical four chamber (A4C) and parasternal long axis (PLAX) views for PAH diagnosis. To assess the importance of different heart chambers in PAH diagnosis, we developed a novel Chamber Attention Module. Results:: The experimental results demonstrated that: 1) The substantial correspondence between our obtained chamber attention vector and clinical expertise suggested that our model was highly interpretable, potentially uncovering diagnostic insights overlooked by the clinical community. 2) The proposed CAN model exhibited superior image-level accuracy and faster convergence on the internal validation dataset compared to the other four models. Furthermore, our CAN model outperformed the others on the external test dataset, with image-level accuracies of 82.53% and 83.32% for A4C and PLAX, respectively. 3) Implementation of the voting strategy notably enhanced the positive predictive value (PPV) and negative predictive value (NPV) of individual-level classification results, enhancing the reliability of our classification outcomes. Conclusions:: These findings indicate that CAN is a feasible technique for AI-assisted PAH diagnosis, providing new insights into cardiac structural changes observed in echocardiography.

Switchgear is a vital component of modern power systems, responsible for regulating the flow of electrical power. Contact issues, irregular loads, and other similar problems can cause switchgear to overheat, leading to unexpected disturbances and potential damage to the power equipment. Thermal imaging shows significant potential and is increasingly employed to detect faults in power equipment. However, the unique characteristics of thermal images often pose challenges to accurate fault detection. This research aims to study the effectiveness of transfer learning architectures for switchgear fault detection. This paper applies eleven transfer learning architectures, including SqueezeNet, GoogLeNet, InceptionV3, DenseNet201, ResNet50, Xception, InceptionResNetV2, ShuffleNet, EfficientNetB0, AlexNet, and VGG19. The results of the testing phase demonstrated that the application of transfer learning by fine-tuning pre-trained convolutional neural network architectures was highly effective in the classification of thermal images captured from switchgear units. The models achieved accuracy rates between 83.87% and 98.38%, and values of F1-Scores between 83.11% and 98.34% in the pre-trained architectures.

Internet of Things (IoT) sensor networks are connected systems of physical devices set with actuators, sensors, and communication abilities, allowing them to gather, spread, and exchange information with centralized methods. These networks are essential in numerous businesses, such as healthcare, manufacturing, agriculture, and smart cities, as they deliver real-time observation, data-driven insights, and automation. Threat recognition in IoT sensor networks is a vital feature of safeguarding the protection and consistency of interconnected systems in the IoT. As IoT sensor networks endure to increase across various industries, the vulnerability to malicious actions and cyber-attacks increases. Threat recognition utilizing deep learning (DL) leverages neural networks to examine complex patterns and anomalies in data, permitting the identification of potential safety threats. DL techniques like convolutional neural networks (CNNs) or recurrent neural networks (RNNs) excel at learning complex representations of data and feature extraction, making them suitable for identifying sophisticated attacks in different fields, including cybersecurity. This research develops a Hybrid Sine-Cosine Chimp Optimization Feature Selection with a Deep Learning (HSCCOFS-DL) approach for Threat Recognition in IoT Sensor Networks. The foremost aim of the HSCCOFS-DL system lies in the automated detection of threats using DL models. To accomplish this, the HSCCOFS-DL approach undergoes a data normalization process. Besides, the selection of features can be performed using the HSCCO algorithm. Meanwhile, the symmetrical autoencoder (SAE) technique effectively classifies threats. Finally, the sparrow search algorithm (SSA) can be applied to the selection of the hyperparameter of the SAE system. The experimental assessment of the HSCCOFS-DL technique takes place on a benchmark dataset. The simulation results indicated that the HSCCOFS-DL approach attains enhanced performance over other methods.

This research presents a novel approach to robotic grasping by integrating an attention mechanism and advanced U-Net architectures, specifically UNet and UNet++, into the Generative Grasping Convolutional Neural Network (GG-CNN). The proposed method, Attention UNet++ GG-CNN, aims to improve the performance and accuracy of robotic grasping predictions significantly. The attention mechanism allows the model to focus on the most relevant features in the depth image, enhancing the quality of grasp predictions. The UNet and UNet++ structures, renowned for their efficiency in semantic segmentation tasks, are adapted to generate a pixel-wise grasp quality map, providing a one-to-one mapping from depth images. The UNet++ structure, with its nested and dense skip pathways, further improves the model's ability to capture and propagate low-level details to higher-level features. This approach overcomes the limitations of traditional deep learning grasping techniques by avoiding discrete sampling of grasp candidates and reducing computation times. Preliminary results indicate that the Attention UNet++ GG-CNN significantly improves the accuracy and performance of robotic grasping,with the Attention UNet++ GG-CNN achieving Intersection over Union (IoU) scores up to 95.86%, and average performance metrics across 25%, 50%, and 75% thresholds reaching as high as 88.29%, confirming its potential for more effective and reliable robotic manipulation in various settings, paving the way for more efficient and reliable robotic manipulation in dynamic environments.

In this paper, we propose mesoscience-guided deep learning (MGDL), a deep learning modeling approach guided by mesoscience, to study complex systems. When establishing sample dataset based on the same system evolution data, different from the operation of conventional deep learning method, MGDL introduces the treatment of the dominant mechanisms of complex system and interactions between them according to the principle of compromise in competition (CIC) in mesoscience. Mesoscience constraints are then integrated into the loss function to guide the deep learning training. Two methods are proposed for the addition of mesoscience constraints. The physical interpretability of the model-training process is improved by MGDL because guidance and constraints based on physical principles are provided. MGDL was evaluated using a bubbling bed modeling case and compared with traditional techniques. With a much smaller training dataset, the results indicate that mesoscience-constraint-based model training has distinct advantages in terms of convergence stability and prediction accuracy, and it can be widely applied to various neural network configurations. The MGDL approach proposed in this paper is a novel method for utilizing the physical background information during deep learning model training. Further exploration of MGDL will be continued in the future.

In this paper, we propose deep learning-based channel estimation and pilot reduction for mmWave point-to-point multi-input multi-output systems. The proposed scheme consists of a two-step approach where the first step is applying a denoising autoencoder for channel estimation. With the denoising characteristic of autoencoder, sparse channel estimation can be conducted although the orthogonality of pilot sequences is not guaranteed due to shorter pilots. The second step is exploiting the temporal correlation of the channel, using the previous estimate to extract information for the current estimate. Through simulation, the proposed scheme shows superior performance with reduced pilots.

Determining homogeneous domains statistically is helpful for engineering geological modeling and rock mass stability evaluation. In this text, a technique that can integrate lithology, geotechnical and structural information is proposed to delineate homogeneous domains. This technique is then applied to a high and steep slope along a road. First, geological and geotechnical domains were described based on lithology, faults, and shear zones. Next, topological manifolds were used to eliminate the incompatibility between orientations and other parameters (i.e. trace length and roughness) so that the data concerning various properties of each discontinuity can be matched and characterized in the same Euclidean space. Thus, the influence of implicit combined effect in between parameter sequences on the homogeneous domains could be considered. Deep learning technique was employed to quantify abstract features of the characterization images of discontinuity properties, and to assess the similarity of rock mass structures. The results show that the technique can effectively distinguish structural variations and outperform conventional methods. It can handle multisource engineering geological information and multiple discontinuity parameters. This technique can also minimize the interference of human factors and delineate homogeneous domains based on orientations or multi-parameter with arbitrary distributions to satisfy different engineering requirements.

Data-driven approaches such as neural networks are increasingly used for deep excavations due to the growing amount of available monitoring data in practical projects. However, most neural network models only use the data from a single monitoring point and neglect the spatial relationships between multiple monitoring points. Besides, most models lack flexibility in providing predictions for multiple days after monitoring activity. This study proposes a sequence-to-sequence (seq2seq) two-dimensional (2D) convolutional long short-term memory neural network (S2SCL2D) for predicting the spatiotemporal wall deflections induced by deep excavations. The model utilizes the data from all monitoring points on the entire wall and extracts spatiotemporal features from data by combining the 2D convolutional layers and long short-term memory (LSTM) layers. The S2SCL2D model achieves a long-term prediction of wall deflections through a recursive seq2seq structure. The excavation depth, which has a significant impact on wall deflections, is also considered using a feature fusion method. An excavation project in Hangzhou, China, is used to illustrate the proposed model. The results demonstrate that the S2SCL2D model has superior prediction accuracy and robustness than that of the LSTM and S2SCL1D (one-dimensional) models. The prediction model demonstrates a strong generalizability when applied to an adjacent excavation. Based on the long-term prediction results, practitioners can plan and allocate resources in advance to address the potential engineering issues.

This work proposes a method for automatic standardized assessment of bone marrow volume and spatial distribution of the proton density fat fraction (PDFF) in vertebral bodies. Intra- and interindividual variability in size and shape of vertebral bodies is a challenge for comparable interindividual evaluation and monitoring of changes in the composition and distribution of bone marrow due to aging and/or intervention. Based on deep learning image segmentation, bone marrow PDFF of single vertebral bodies is mapped to a cylindrical template and corrected for the inclination with respect to the horizontal plane. The proposed technique was applied and tested in a cohort of 60 healthy (30 males, 30 females) individuals. Obtained bone marrow volumes and mean PDFF values are comparable to former manual and (semi-)automatic approaches. Moreover, the proposed method allows shape-independent characterization of the spatial PDFF distribution inside vertebral bodies.

Owing to their limited accuracy and narrow applicability, current antimicrobial peptide (AMP) prediction models face obstacles in industrial application. To address these limitations, we developed and improved an AMP prediction model using Comparing and Optimizing Multiple DEep Learning (COMDEL) algorithms, coupled with high-throughput AMP screening method, finally reaching an accuracy of 94.8% in test and 88% in experiment verification, surpassing other state-of-the-art models. In conjunction with COMDEL, we employed the phage-assisted evolution method to screen Sortase in vivo and developed a cell-free AMP synthesis system in vitro, ultimately increasing AMPs yields to a range of 0.5–2.1 g/L within hours. Moreover, by multi-omics analysis using COMDEL, we identified Lactobacillus plantarum as the most promising candidate for AMP generation among 35 edible probiotics. Following this, we developed a microdroplet sorting approach and successfully screened three L. plantarum mutants, each showing a twofold increase in antimicrobial ability, underscoring their substantial industrial application values.

COVID-19 detection is mainly based on molecular testing, especially polymerase chain reaction (PCR) analyses, to find the occurrence of the virus's genetic material in respiratory samples. Computed tomography (CT) scans and X-rays are the two chest imaging techniques, which may provide more details on lungs, aiding in the diagnosis. Radiological findings in COVID-19 cases typically involve special features such as bilateral involvement, ground-glass opacities, and consolidation of the lungs. Deep learning (DL) methods, particularly convolutional neural networks (CNNs) are used for analyzing COVID-19 infection. By training these techniques on large datasets of annotated chest X-rays (CXRs), researches have attained superior outcomes in precisely detecting the possible COVID-19 cases. Leveraging artificial intelligence together with radiological expertise holds great potential for facilitating prompt intervention in the ongoing global efforts and improving the efficiency of early detection. This article focuses on the design of mother optimization algorithm with DL-enabled detection and classification (MOA-DLEDC) technique for COVID-19 diagnosis using CXR images. In the MOA-DLEDC technique, the adaptive median filtering (AMF) approach is used to eliminate the existence of noise. In addition, the complex and intrinsic feature patterns are derived from the Efficient DenseNet model. Besides, the hyperparameter optimization of the Efficient DenseNet model takes place using MOA. For the COVID-19 detection process, the MOA-DLEDC technique applies sparse autoencoder (SAE) model. A detailed set of experiments was conducted on the CXR dataset to highlight the promising results of the MOA-DLEDC technique. The extensive results inferred that the used technique gains superior performance over other models with maximum accuracy of 99.59 %.

Sliding bearings are crucial rotating mechanical components in nuclear power plants, and their failures can result in severe economic losses and human casualties. Deep learning provides a new approach to bearing fault diagnosis, but there is currently a lack of a universal fault diagnosis model for studying bearing-rotor systems under various operating conditions, speeds and faults. Research on bearing-rotor systems supported by sliding bearings is limited, leading to insufficient fault data. To address these issues, this paper proposes a fault diagnosis model framework for bearing-rotor systems based on a deep convolutional generative adversarial network (TF-DLGAN). This model not only exhibits outstanding fault diagnosis performance but also addresses the issue of insufficient fault data. An experimental platform is constructed to conduct fault experiments under various operating conditions, speeds and faults, establishing a dataset for sliding bearing-rotor system faults. Finally, the model's effectiveness is validated using this dataset.

Background: Uncorrected refractive error is a major cause of vision impairment worldwide and its increasing prevalent necessitates effective screening and management strategies. Meanwhile, deep learning, a subset of Artificial Intelligence, has significantly advanced ophthalmological diagnostics by automating tasks that required extensive clinical expertise. Although recent studies have investigated the use of deep learning models for refractive power detection through various imaging techniques, a comprehensive systematic review on this topic is has yet be done. This review aims to summarise and evaluate the performance of ocular image-based deep learning models in predicting refractive errors. Main text: We search on three databases (PubMed, Scopus, Web of Science) up till June 2023, focusing on deep learning applications in detecting refractive error from ocular images. We included studies that had reported refractive error outcomes, regardless of publication years. We systematically extracted and evaluated the continuous outcomes (sphere, SE, cylinder) and categorical outcomes (myopia), ground truth measurements, ocular imaging modalities, deep learning models, and performance metrics, adhering to PRISMA guidelines. Nine studies were identified and categorised into three groups: retinal photo-based (n ​= ​5), OCT-based (n ​= ​1), and external ocular photo-based (n ​= ​3).For high myopia prediction, retinal photo-based models achieved AUC between 0.91 and 0.98, sensitivity levels between 85.10% and 97.80%, and specificity levels between 76.40% and 94.50%. For continuous prediction, retinal photo-based models reported MAE ranging from 0.31D to 2.19D, and R2 between 0.05 and 0.96. The OCT-based model achieved an AUC of 0.79–0.81, sensitivity of 82.30% and 87.20% and specificity of 61.70%–68.90%. For external ocular photo-based models, the AUC ranged from 0.91 to 0.99, sensitivity of 81.13%–84.00% and specificity of 74.00%–86.42%, MAE ranges from 0.07D to 0.18D and accuracy ranges from 81.60% to 96.70%. The reported papers collectively showed promising performances, in particular the retinal photo-based and external eye photo -based DL models. Conclusions: The integration of deep learning model and ocular imaging for refractive error detection appear promising. However, their real-world clinical utility in current screening workflow have yet been evaluated and would require thoughtful consideration in design and implementation.

The Advanced Geosynchronous Radiation Imager (AGRI) is a mission-critical instrument for the Fengyun series of satellites. AGRI acquires full-disk images every 15 min and views East Asia every 5 min through 14 spectral bands, enabling the detection of highly variable aerosol optical depth (AOD). Quantitative retrieval of AOD has hitherto been challenging, especially over land. In this study, an AOD retrieval algorithm is proposed that combines deep learning and transfer learning. The algorithm uses core concepts from both the Dark Target (DT) and Deep Blue (DB) algorithms to select features for the machine-learning (ML) algorithm, allowing for AOD retrieval at 550 nm over both dark and bright surfaces. The algorithm consists of two steps: ① A baseline deep neural network (DNN) with skip connections is developed using 10 min Advanced Himawari Imager (AHI) AODs as the target variable, and ② sunphotometer AODs from 89 ground-based stations are used to fine-tune the DNN parameters. Out-of-station validation shows that the retrieved AOD attains high accuracy, characterized by a coefficient of determination (R2) of 0.70, a mean bias error (MBE) of 0.03, and a percentage of data within the expected error (EE) of 70.7%. A sensitivity study reveals that the top-of-atmosphere reflectance at 650 and 470 nm, as well as the surface reflectance at 650 nm, are the two largest sources of uncertainty impacting the retrieval. In a case study of monitoring an extreme aerosol event, the AGRI AOD is found to be able to capture the detailed temporal evolution of the event. This work demonstrates the superiority of the transfer-learning technique in satellite AOD retrievals and the applicability of the retrieved AGRI AOD in monitoring extreme pollution events.

Introduction: The risk of complications associated with transvenous ICDs make the subcutaneous implantable cardiac defibrillator (S-ICD) a valuable alternative in patients with adult congenital heart disease (ACHD). However, higher S-ICD ineligibility and higher inappropriate shock rates-mostly caused by T wave oversensing (TWO)- are observed in this population. We report a novel application of deep learning methods to screen patients for S-ICD eligibility over a longer period than conventional screening. Methods: Adult patients with ACHD and a control group of normal subjects were fitted with a 24-h Holters to record their S-ICD vectors. Their T:R ratio was analysed utilising phase space reconstruction matrices and a deep learning-based model to provide an in-depth description of the T: R variation plot for each vector. T: R variation was compared statistically using t-test. Results: 13 patients (age 37.4 ± 7.89 years, 61.5 % male, 6 ACHD and 7 control subjects) were enrolled. A significant difference was observed in the mean and median T: R values between the two groups (p

ABSTRACT: Identifying genome-enabled methods that provide more accurate genomic prediction is crucial when evaluating complex traits such as dairy cow behavior. In this study, we aimed to compare the predictive performance of traditional genomic prediction methods and deep learning algorithms for genomic prediction of milking refusals (MREF) and milking failures (MFAIL) in North American Holstein cows measured by automatic milking systems (milking robots). A total of 1,993,509 daily records from 4,511 genotyped Holstein cows were collected by 36 milking robot stations. After quality control, 57,600 SNPs were available for the analyses. Four genomic prediction methods were considered: Bayesian least absolute shrinkage and selection operator (LASSO), multiple layer perceptron (MLP), convolutional neural network (CNN), and GBLUP. We implemented the first 3 methods using the Keras and TensorFlow libraries in Python (v.3.9) but the GBLUP method was implemented using the BLUPF90+ family programs. The accuracy of genomic prediction (mean square error) for MREF and MFAIL was 0.34 (0.08) and 0.27 (0.08) based on LASSO, 0.36 (0.09) and 0.32 (0.09) for MLP, 0.37 (0.08) and 0.30 (0.09) for CNN, and 0.35 (0.09) and 0.31(0.09) based on GBLUP, respectively. Additionally, we observed a lower reranking of top selected individuals based on the MLP versus CNN methods compared with the other approaches for both MREF and MFAIL. Although the deep learning methods showed slightly higher accuracies than GBLUP, the results may not be sufficient to justify their use over traditional methods due to their higher computational demand and the difficulty of performing genomic prediction for nongenotyped individuals using deep learning procedures. Overall, this study provides insights into the potential feasibility of using deep learning methods to enhance genomic prediction accuracy for behavioral traits in livestock. Further research is needed to determine their practical applicability to large dairy cattle breeding programs.

Complex mapping relationships between high-entropy alloy compositions and performances struggle to precisely elucidate by traditional machine learning models, hindering the accurate and efficient design of high-performance alloys. In this work, a novel alloy design strategy combined self-decision deep neural network, data reconstruction with network structure optimization was proposed, which effectively enhanced the model's ability to predict the mapping relationships between high-dimensional composition spaces and mechanical properties of alloy. Compared to other machine learning methods, significantly improves model prediction accuracy (hardness model: 97.98%; strength model: 94.40%). Through above model, the mechanical properties of 308 new (CuNiMn)-X high-entropy copper alloys was studied added other elements such as Al, Ti, Cr, and Fe, which designed and produced a novel (CuNiMn)60Al20Cr20 alloy with high hardness of 474 HV, high compressive strength of 2322 MPa, and high fracture strain of 25.20%. The primary factors influencing alloy performance and their respective thresholds were also quantitatively revealed (hardness: electronegativity deviation (0.02, 0.1) and nuclear charge deviation (0, 0.25); compressive strength: mean nuclear charge (25, 30) and mean atomic radius (134 p.m.)). These findings provide a new perspective and approach for the design and mechanism understanding of high-performance alloys with complex compositions.

Liquid metal heat pipes play a critical role in various high-temperature applications, with their optimization being pivotal to achieving optimal thermal performance. In this study, a deep learning based genetic algorithm is suggested to optimize the operating conditions of liquid metal heat pipes. The optimization performance was investigated in both single and multi-variable optimization schemes, considering the operating conditions of heat load, inclination angle, and filling ratio. The single-variable optimization indicated reasonable performance for various conditions, reinforcing the potential applicability of the optimization method across a broad spectrum of high-temperature industries. The multi-variable optimization revealed an almost congruent performance level to single-variable optimization, suggesting that the robustness of optimization method is not compromised with additional variables. Furthermore, the generalization performance of the optimization method was investigated by conducting an experimental investigation, proving a similar performance. This study underlines the potential of optimizing the operating condition of heat pipes, with significant consequences in sectors such as high temperature field, thereby offering a pathway to more efficient, cost-effective thermal solutions.

Manual segmentation of brain tumors using structural magnetic resonance imaging (MRI) is an arduous and time-consuming task. Therefore, automatic and robust segmentation will considerably influence neuro-oncological clinical trials by reducing excessive manual annotation time. Herein, we propose a deep learning model that automatically segments brain tumors even in cases of missing MRI sequences, which are common in practical clinical settings. To address this issue, we enhance a generative adversarial network (GAN) by incorporating a squeeze-and-excitation (SE) attention module into its generator and a PatchGAN into its discriminator. The SE module recalibrates channel responses by explicitly modeling interdependencies, enabling the network to focus on critical regions such as tumor areas. Our proposed generative model is optimized using a combination of adversarial, structural similarity, and mean absolute error losses to synthesize missing MRI sequences more effectively. This enhancement allows our model to synthesize the missing MRI sequence (fluid attenuated inversion recovery [FLAIR]) by leveraging information from other available sequences (T1-weighted, T2-weighted, or contrast-enhanced T1-weighted [T1ce]). For the segmentation task, we employ an optimized nnU-Net model, which is trained using existing sequences and evaluated using both available and synthesized sequences (including missing ones), mimicking real-world scenarios where often only limited MRI sequences are available or usable. Our findings reveal a notable enhancement in brain tumor segmentation, as indicated by a significant increase in overall the Dice similarity coefficient (DSC) from 0.688% (when FLAIR is missing) to 0.873% (when using synthesized FLAIR derived from T2). This improvement brings the segmentation performance closer to what was achieved when real FLAIR was available, where the DSC reaches 0.901%. Moreover, our synthesizing model was also tested on two additional datasets: the BraTS 2020 validation set and BraTS Africa 2023 training set, which produces results comparable to those of BraTS 2021, thereby proving its robustness and generalizability. In addition, the resulting tumor segmentations are subsequently employed to assess the response to treatment in cases where all sequences were available and when synthesis was employed, according to response assessment in neuro-oncology criteria.

In the future of healthcare, Blockchain (BC) technology holds immense potential for improving the security and privacy of data. By allowing the secure and immutable storage of medical files and healthcare-related transactions, BC ensured that sensitive medical data remains tamper-proof and open only to authorized parties. Patients have greater control over their data's development, revoking or granting access as required, but healthcare workers can streamline data sharing and ensure the integrity of important data. The decentralized nature of BC networks decreases the risk of centralized data breaches, eventually fostering trust and transparency in the healthcare ecosystems. Conversely, deep learning (DL) has great to revolutionize healthcare diagnostics in the future, offering quick and extremely accurate estimates of medical conditions. This technology has greatly enhanced patient solutions, decreased medical expenses, and improved the burden on medical staff by providing appreciated insights into an extensive range of conditions, from cancer to neurological disorders. With this stimulus, this study presents a novel BC with optimal DL-based secure data sharing and classification (BCODL-SDSC) technique in the future healthcare system. The goal of the BCODL-SDSC technique is to secure and thoroughly examine healthcare data using BC and DL techniques. Primarily, the BCODL-SDSC technique enables BC technology to store and maintain the patient’s data from the procedure of several transactions and enable access control to the various stakeholders. For the security of the medical images, the BCODL-SDSC technique applies the Fractional Order Lorenz system (FOLS) based encryption technique with tuna swarm optimization (TSO) algorithm based optimal key generation process. Finally, a multi-stage process performs the classification of the medical images: MobileNetv1 feature extractor, artificial rabbit’s optimization (ARO) based hyperparameter tuning, and stacked recurrent neural network (SRNN) based classification. The experimental outcome of the BCODL-SDSC technique was examined on a benchmark medical image database. An extensive comparative study reported that the BCODL-SDSC technique reaches an effective performance with other models with a maximum accuracy of 99.11%.

Utilizing Wireless Sensor Networks (WSNs), Internet of Things (IoTs) sensors, and Unmanned Aerial Vehicles (UAVs), in conjunction with optimization techniques and machine learning algorithms, can present a novel approach to Precision Agricultural (PA) crops. Agriculture has transitioned from traditional legacy systems to incorporate advanced smart technologies. Several complex farming problems are utilized to promote a variety of UAV sensing and Artificial Intelligence (AI) algorithms in PA applications of smart agriculture, especially in developing countries. This paper proposes the design of a conceptual UAV sensing system for crop management using a novel classification framework. UAV-based image datasets can be implemented and expanded to various types of crops. In addition, it can classify various types of rice species and detect weed farms as it consists of a multistage process for identifying and monitoring different crops. The proposed classification framework is evaluated using three different real UAV image datasets compared with Naive Bayes, Decision Tree (DT), Bagging, and Random Forest (RF) techniques. The classification performance metrics obtained are as follows: i) the WeedNet dataset with 100 % F1-score, 100 % recall, 100 % precision, and 100 % accuracy; ii) the Rice Seedling dataset with 99.5 % F1-score, 99.5 % recall, 99.5 % precision, and 99.5 % accuracy; and iii) the Rice Varieties dataset with 97.99 % F1-score, 97.99 % recall, 98.14 % precision, and 97.99 % accuracy. The success rates achieved by the proposed classification framework outperform those of other recent state-of-the-art techniques, demonstrating its efficacy in crop management applications.

Smart ports, utilizing advanced and hybrid technologies, are gaining increasing attention for application in the maritime industry, with driver assistance and autonomous driving being pivotal in container-terminal operations. This study introduces a novel approach for enhancing object detection and distance estimation, focusing principally on decision support for reach stacker container handlers in port terminals by integrating generative and deep learning models. The EfficientDet model, enriched with integrated k-means clustering, is developed to detect and classify objects using a practical dataset of labeled images based on visual features. Moreover, generative models, specifically the diffusion model and generative adversarial network, are utilized to generate depth scenes for estimating object distances. Experimental results indicate that the proposed approach yields superior object detection and distance estimation outcomes in port terminal operations, characterized by high accuracy and reduced computational cost. The proposed method exhibits potential for application across various industries, including transportation, logistics, and security, where precise object detection and distance estimation are vital for efficient and secure operations.

Leveraging deep learning (DL) to inverse problems has proven transformative in predicting financial futures, mainly in stock price prediction. In terms of financial markets, where predicting stock price is a challenging inverse problem, DL methods like long short-term memory networks (LSTM) and recurrent neural networks (RNN) demonstrate notable prowess. These techniques successfully capture complex patterns and dependencies in historical stock price information, learning to discern subtle correlations and fundamental market trends. DL facilitates the extraction of nuanced features by leveraging large datasets and sophisticated techniques, enabling accurate prediction of future stock prices. This intersection of advanced ML approaches and financial predicting signifies a paradigm shift in leveraging technology to navigate the complexity of the financial market and improves decision-making for traders and investors. This manuscript introduces a Dung Beetle Optimization with a Deep Learning Approach for Solving Inverse Problems in Predicting Financial Futures (DBODL-SIPPFF) technique. The DBODL-SIPPFF technique resolves the inverse problem and predicts the stock prices adequately. In addition, the DBODL-SIPPFF technique exploits data consistency (DC) optimizer to solve inverse problems. Primarily, linear scaling normalization (LSN) is applied for data normalization. The honey badger algorithm (HBA) is utilized for the feature selection. Moreover, convolutional long short-term memory (ConvLSTM) is used for stock price prediction. Lastly, the DBO technique is utilized for the optimum hyperparameter selection of the ConvLSTM model. The empirical assessment of the DBODL-SIPPFF technique takes place under diverse stock datasets. The obtained values stated that the DBODL-SIPPFF technique performs optimally over compared methods.

Facial Expression Recognition (FER) categorizes various human emotions by analyzing the features of the face, so it plays a vital role in recognizing emotions. Prior studies have focused on the issue of recognizing emotions through voices or speech. Addressing the existing method issues, this approach aims to detect voices and three-dimensional images using appropriate datasets and novel deep-learning techniques. In this research, the valid Audio-Visual datasets Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS), Acted Facial Expression in the Wild (AFEW), and eNTERFACE’05 datasets are chosen for analysis. RAVDESS dataset contains audio, AFEW, and eNTERFACE and has three-dimensional images of humans, i.e., 3D images. SMOTE technique is presented for solving overfitting problems to balance the dataset by oversampling and under-sampling process. The research employs the Federated 3D-CNN technique to predict the accurate emotions of humans. The 3D Convolutional Neural Network (3DCNN) predicts accurate information of a person at any angle in image processing. Mel Frequency Cepstrum Coefficient (MFCC) is used to extract and fine-tune the voices. A significant contribution of Federated Learning with 3D-Convolutional Neural Network is executed for multiple clients at a time through global and local updates of weights. The proposed framework achieves a prediction accuracy of 95.72 % when compared with existing methods. This approach helps in many applications, such as analyzing emotions, healthcare, etc.

Video anomaly detection is a very important research topic, especially in the field of automatic driving, and timely acquisition of road conditions ahead is very important for safe driving. Past methods used convolution and other operations, resulting in severe delays or low detection accuracy. This paper introduces a cutting-edge video anomaly detection method that combines the advantages of image quality assessment and attention mechanisms while using image signal processing techniques to enhance the data. Our approach first uses image quality metrics to evaluate visual fidelity anomalies for each video frame, providing a computationally simple and effective method for initial anomaly recognition. Then, an attention mechanism is integrated into two deep learning models including convolutional neural networks (CNNs) and long short-term memory (LSTM) networks. This integration enables the system to dynamically focus on significant regions in the video. This capability is unmatched by traditional analysis methods. The dual-structure system shows higher sensitivity to various anomalies. In addition, the system is also equipped with image signals, which makes the data feature extraction more precise. Through thorough testing of the benchmark dataset, the proposed approach has been shown to significantly outperform current state-of-the-art models. It especially excels in challenging monitoring scenarios. For example, in the UCSD dataset, we achieved an area under curve (AUC) value of 99.9%, which exceeds most current methods.

The 4,6-substituted-1,3,5-triazin-2(1H)-ones are promising inhibitors of human DNA topoisomerase IIα. To further develop this chemical class targeting the enzyme´s ATP binding site, the triazin-2(1H)-one substitution position 6 was optimized. Inspired by binding of preclinical substituted 9H-purine derivative, bicyclic substituents were incorporated at position 6 and the utility of this modification was validated by a combination of molecular simulations, dynamic pharmacophores, and free energy calculations. Considering also predictions of Deepfrag, a software developed for structure-based lead optimization based on deep learning, compounds with both bicyclic and monocyclic substitutions were synthesized and investigated for their inhibitory activity. The SAR data showed that the bicyclic substituted compounds exhibited good inhibition of topo IIα, comparable to their mono-substituted counterparts. Further evaluation on a panel of human protein kinases showed selectivity for the inhibition of topo IIα. Mechanistic studies indicated that the compounds acted predominantly as catalytic inhibitors, with some exhibiting topo IIα poison effects at higher concentrations. Integration of STD NMR experiments and molecular simulations, provided insights into the binding model and highlighted the importance of the Asn120 interaction and hydrophobic interactions with substituents at positions 4 and 6. In addition, NCI-60 screening demonstrated cytotoxicity of the compounds with bicyclic substituents and identified sensitive human cancer cell lines, underlining the translational relevance of our findings for further preclinical development of this class of compounds. The study highlights the synergy between simulation and AI-based approaches in efficiently guiding molecular design for drug optimization, which has implications for further preclinical development of this class of compounds.

Incorporating ground-breaking technologies such as deep learning (DL) has revolutionized predictive modelling in the rapidly evolving landscape of the finance sector. DL approaches, capable of extracting complex patterns from vast data collections, become an efficient approach for predicting financial trends. By integrating the complex neural network architecture with comprehensive datasets, including investor sentiment, market indicators, and economic variables, finance experts have introduced prediction models well known for their ability to capture the nuanced dynamics of financial markets with remarkable performance. Incorporating DL approaches within the finance sector provided the basis for more informed decision-making, enabling institutions, investors, and analysts to capitalize on emerging opportunities with greater confidence and precision and navigate market volatility. This study develops a novel quasi-oppositional coot algorithm with a deep learning-based predictive method on the financial sector (QOCODL-PMFC) technique. The QOCODL-PMFC technique aims to perform a prediction process on the financial sector. The QOCODL-PMFC method applies min-max normalization to measure the input dataset into a meaningful format to achieve this. Next, the QOCODL-PMFC method designs the QOCO technique for selecting an optimal set of features. The QOCODL-PMFC technique applies the attention bidirectional gated recurrent unit (ABiGRU) model for the prediction process. The Harris Hawks Optimization (HHO) model is utilized to boost the performance of the ABiGRU network. The simulation evaluation of the QOCODL-PMFC technique is tested under a benchmark finance dataset. The experimental values of the QOCODL-PMFC technique exhibit a minimal MSE of 0.7452 over other models.

Artificial Intelligence is transforming drug discovery, particularly in the hit identification phase of therapeutic compounds. One tool that has been instrumental in this transformation is Quantitative Structure-Activity Relationship (QSAR) analysis. This computer-aided drug design tool uses machine learning to predict the biological activity of new compounds based on the numerical representation of chemical structures against various biological targets. With diabetes mellitus becoming a significant health challenge in recent times, there is intense research interest in modulating antidiabetic drug targets. α-Glucosidase is an antidiabetic target that has gained attention due to its ability to suppress postprandial hyperglycaemia, a key contributor to diabetic complications. This review explored a detailed approach to developing QSAR models, focusing on strategies for generating input variables (molecular descriptors) and computational approaches ranging from classical machine learning algorithms to modern deep learning algorithms. We also highlighted studies that have used these approaches to develop predictive models for α-glucosidase inhibitors to modulate this critical antidiabetic drug target.

In the field of computational oncology, patient status is often assessed using radiology-genomics, which includes two key technologies and data, such as radiology and genomics. Recent advances in deep learning have facilitated the integration of radiology-genomics data, and even new omics data, significantly improving the robustness and accuracy of clinical predictions. These factors are driving artificial intelligence (AI) closer to practical clinical applications. In particular, deep learning models are crucial in identifying new radiology-genomics biomarkers and therapeutic targets, supported by explainable AI (xAI) methods. This review focuses on recent developments in deep learning for radiology-genomics integration, highlights current challenges, and outlines some research directions for multimodal integration and biomarker discovery of radiology-genomics or radiology-omics that are urgently needed in computational oncology.

Therapeutic antibodies are an important class of biopharmaceuticals. With the rapid development of deep learning methods and the increasing amount of antibody data, antibody generative models have made great progress recently. They aim to solve the antibody space searching problems and are widely incorporated into the antibody development process. Therefore, a comprehensive introduction to the development methods in this field is imperative. Here, we collected 34 representative antibody generative models published recently and all generative models can be divided into three categories: sequence-generating models, structure-generating models, and hybrid models, based on their principles and algorithms. We further studied their performance and contributions to antibody sequence prediction, structure optimization, and affinity enhancement. Our manuscript will provide a comprehensive overview of the status of antibody generative models and also offer guidance for selecting different approaches.

Cathepsin L (CTSL) is a promising therapeutic target for metabolic disorders. Current pharmacological interventions targeting CTSL have demonstrated potential in reducing body weight gain, serum insulin levels, and improving glucose tolerance. However, the clinical application of CTSL inhibitors remains limited. In this study, we used a combination of artificial intelligence and experimental methods to identify new CTSL inhibitors from natural products. Through a robust deep learning model and molecular docking, we screened 150 molecules from natural products for experimental validation. At a concentration of 100 µM, we found that 36 of them exhibited more than 50 % inhibition of CTSL. Notably, 13 molecules displayed over 90 % inhibition and exhibiting concentration-dependent effects. The molecular dynamics simulation on the two most potent inhibitors, Plumbagin and Beta-Lapachone, demonstrated stable interaction at the CTSL active site. Enzyme kinetics studies have shown that these inhibitors exert an uncompetitive inhibitory effect on CTSL. In conclusion, our research identifies Plumbagin and Beta-Lapachone as potential CTSL inhibitors, offering promising candidates for the treatment of metabolic disorders and illustrating the effectiveness of artificial intelligence in drug discovery.

Background: Bone marrow adipose tissue (BMAT) represents > 10% fat mass in healthy humans and can be measured by magnetic resonance imaging (MRI) as the bone marrow fat fraction (BMFF). Human MRI studies have identified several diseases associated with BMFF but have been relatively small scale. Population-scale studies therefore have huge potential to reveal BMAT’s true clinical relevance. The UK Biobank (UKBB) is undertaking MRI of 100,000 participants, providing the ideal opportunity for such advances. Objective: To establish deep learning for high-throughput multi-site BMFF analysis from UKBB MRI data. Materials and methods: We studied males and females aged 60–69. Bone marrow (BM) segmentation was automated using a new lightweight attention-based 3D U-Net convolutional neural network that improved segmentation of small structures from large volumetric data. Using manual segmentations from 61–64 subjects, the models were trained to segment four BM regions of interest: the spine (thoracic and lumbar vertebrae), femoral head, total hip and femoral diaphysis. Models were tested using a further 10–12 datasets per region and validated using datasets from 729 UKBB participants. BMFF was then quantified and pathophysiological characteristics assessed, including site- and sex-dependent differences and the relationships with age, BMI, bone mineral density, peripheral adiposity, and osteoporosis. Results: Model accuracy matched or exceeded that for conventional U-Nets, yielding Dice scores of 91.2% (spine), 94.5% (femoral head), 91.2% (total hip) and 86.6% (femoral diaphysis). One case of severe scoliosis prevented segmentation of the spine, while one case of Non-Hodgkin Lymphoma prevented segmentation of the spine, femoral head and total hip because of T2 signal depletion; however, successful segmentation was not disrupted by any other pathophysiological variables. The resulting BMFF measurements confirmed expected relationships between BMFF and age, sex and bone density, and identified new site- and sex-specific characteristics. Conclusions: We have established a new deep learning method for accurate segmentation of small structures from large volumetric data, allowing high-throughput multi-site BMFF measurement in the UKBB. Our findings reveal new pathophysiological insights, highlighting the potential of BMFF as a novel clinical biomarker. Applying our method across the full UKBB cohort will help to reveal the impact of BMAT on human health and disease.

RNA-binding proteins (RBPs) are key post-transcriptional regulators, and the malfunctions of RBP-RNA binding lead to diverse human diseases. However, prediction of RBP binding sites is largely based on RNA sequence features, whereas in vivo RNA structural features based on high-throughput sequencing are rarely incorporated. Here, we designed a deep bimodal information fusion network called DeepFusion for unraveling protein-RNA interactions by incorporating structural features derived from DMS-seq data. DeepFusion integrates two sub-models to extract local motif-like information and long-term context information. We show that DeepFusion performs best compared with other cutting-edge methods with only sequence inputs on two datasets. DeepFusion’s performance is further improved with bimodal input after adding in vivo DMS-seq structural features. Furthermore, DeepFusion can be used for analyzing RNA degradation, demonstrating significantly different RBP-binding scores in genes with slow degradation rates versus those with rapid degradation rates. DeepFusion thus provides enhanced abilities for further analysis of functional RNAs. DeepFusion’s code and data are available at http://bioinfo.org/deepfusion/.

Gene transcription is an essential process involved in all aspects of cellular functions with significant impact on biological traits and diseases. This process is tightly regulated by multiple elements that co-operate to jointly modulate the transcription levels of target genes. To decipher the complicated regulatory network, we present a novel multi-view attention-based deep neural network that models the relationship between genetic, epigenetic, and transcriptional patterns and identifies co-operative regulatory elements (COREs). We applied this new method, named DeepCORE, to predict transcriptomes in various tissues and cell lines, which outperformed the state-of-the-art algorithms. Furthermore, DeepCORE contains an interpreter that extracts the attention values embedded in the deep neural network, maps the attended regions to putative regulatory elements, and infers COREs based on correlated attentions. The identified COREs are significantly enriched with known promoters and enhancers. Novel regulatory elements discovered by DeepCORE showed epigenetic signatures consistent with the status of histone modification marks.

With the advent of digital pathology and microscopic systems that can scan and save whole slide histological images automatically, there is a growing trend to use computerized methods to analyze acquired images. Among different histopathological image analysis tasks, nuclei instance segmentation plays a fundamental role in a wide range of clinical and research applications. While many semi- and fully-automatic computerized methods have been proposed for nuclei instance segmentation, deep learning (DL)-based approaches have been shown to deliver the best performances. However, the performance of such approaches usually degrades when tested on unseen datasets.In this work, we propose a novel method to improve the generalization capability of a DL-based automatic segmentation approach. Besides utilizing one of the state-of-the-art DL-based models as a baseline, our method incorporates non-deterministic train time and deterministic test time stain normalization, and ensembling to boost the segmentation performance. We trained the model with one single training set and evaluated its segmentation performance on seven test datasets. Our results show that the proposed method provides up to 4.9%, 5.4%, and 5.9% better average performance in segmenting nuclei based on Dice score, aggregated Jaccard index, and panoptic quality score, respectively, compared to the baseline segmentation model.

Therapeutic antibody development faces challenges due to high viscosities and aggregation tendencies. The spatial charge map (SCM) and spatial aggregation propensity (SAP) are computational techniques that aid in predicting viscosity and aggregation, respectively. These methods rely on structural data derived from molecular dynamics (MD) simulations, which are computationally demanding. DeepSCM, a deep learning surrogate model based on sequence information to predict SCM, was recently developed to screen high-concentration antibody viscosity. This study further utilized a dataset of 20,530 antibody sequences to train a convolutional neural network deep learning surrogate model called Deep Spatial Properties (DeepSP). DeepSP directly predicts SAP and SCM scores in different domains of antibody variable regions based solely on their sequences without performing MD simulations. The linear correlation coefficient between DeepSP scores and MD-derived scores for 30 properties achieved values between 0.76 and 0.96 with an average of 0.87. DeepSP descriptors were employed as features to build machine learning models to predict the aggregation rate of 21 antibodies, and the performance is similar to the results obtained from the previous study using MD simulations. This result demonstrates that the DeepSP approach significantly reduces the computational time required compared to MD simulations. The DeepSP model enables the rapid generation of 30 structural properties that can also be used as features in other research to train machine learning models for predicting various antibody stability using sequences only. DeepSP is freely available as an online tool via https://deepspwebapp.onrender.com and the codes and parameters are freely available at https://github.com/Lailabcode/DeepSP.

Molecular docking is a widely used technique in drug discovery to predict the binding mode of a given ligand to its target. However, the identification of the near-native binding pose in docking experiments still represents a challenging task as the scoring functions currently employed by docking programs are parametrized to predict the binding affinity, and, therefore, they often fail to correctly identify the ligand native binding conformation. Selecting the correct binding mode is crucial to obtaining meaningful results and to conveniently optimizing new hit compounds. Deep learning (DL) algorithms have been an area of a growing interest in this sense for their capability to extract the relevant information directly from the protein-ligand structure. Our review aims to present the recent advances regarding the development of DL-based pose selection approaches, discussing limitations and possible future directions. Moreover, a comparison between the performances of some classical scoring functions and DL-based methods concerning their ability to select the correct binding mode is reported. In this regard, two novel DL-based pose selectors developed by us are presented.

B-cell epitope identification plays a vital role in the development of vaccines, therapies, and diagnostic tools. Currently, molecular docking tools in B-cell epitope prediction are heavily influenced by empirical parameters and require significant computational resources, rendering a great challenge to meet large-scale prediction demands. When predicting epitopes from antigen-antibody complex, current artificial intelligence algorithms cannot accurately implement the prediction due to insufficient protein feature representations, indicating novel algorithm is desperately needed for efficient protein information extraction. In this paper, we introduce a multimodal model called WUREN (Whole-modal Union Representation for Epitope predictioN), which effectively combines sequence, graph, and structural features. It achieved AUC-PR scores of 0.213 and 0.193 on the solved structures and AlphaFold-generated structures, respectively, for the independent test proteins selected from DiscoTope3 benchmark. Our findings indicate that WUREN is an efficient feature extraction model for protein complexes, with the generalizable application potential in the development of protein-based drugs. Moreover, the streamlined framework of WUREN could be readily extended to model similar biomolecules, such as nucleic acids, carbohydrates, and lipids.

Deep learning (DL) has substantially enhanced natural language processing (NLP) in healthcare research. However, the increasing complexity of DL-based NLP necessitates transparent model interpretability, or at least explainability, for reliable decision-making. This work presents a thorough scoping review of explainable and interpretable DL in healthcare NLP. The term “eXplainable and Interpretable Artificial Intelligence” (XIAI) is introduced to distinguish XAI from IAI. Different models are further categorized based on their functionality (model-, input-, output-based) and scope (local, global). Our analysis shows that attention mechanisms are the most prevalent emerging IAI technique. The use of IAI is growing, distinguishing it from XAI. The major challenges identified are that most XIAI does not explore “global” modelling processes, the lack of best practices, and the lack of systematic evaluation and benchmarks. One important opportunity is to use attention mechanisms to enhance multi-modal XIAI for personalized medicine. Additionally, combining DL with causal logic holds promise. Our discussion encourages the integration of XIAI in Large Language Models (LLMs) and domain-specific smaller models. In conclusion, XIAI adoption in healthcare requires dedicated in-house expertise. Collaboration with domain experts, end-users, and policymakers can lead to ready-to-use XIAI methods across NLP and medical tasks. While challenges exist, XIAI techniques offer a valuable foundation for interpretable NLP algorithms in healthcare.

Recent advances in single-cell omics technology have transformed the landscape of cellular and molecular research, enriching the scope and intricacy of cellular characterisation. Perturbation modelling seeks to comprehensively grasp the effects of external influences like disease onset or molecular knock-outs or external stimulants on cellular physiology, specifically on transcription factors, signal transducers, biological pathways, and dynamic cell states. Machine and deep learning tools transform complex perturbational phenomena in algorithmically tractable tasks to formulate predictions based on various types of single-cell datasets. However, the recent surge in tools and datasets makes it challenging for experimental biologists and computational scientists to keep track of the recent advances in this rapidly expanding filed of single-cell modelling. Here, we recapitulate the main objectives of perturbation modelling and summarise novel single-cell perturbation technologies based on genetic manipulation like CRISPR or compounds, spanning across omic modalities. We then concisely review a burgeoning group of computational methods extending from classical statistical inference methodologies to various machine and deep learning architectures like shallow models or autoencoders, to biologically informed approaches based on gene regulatory networks, and to combinatorial efforts reminiscent of ensemble learning. We also discuss the rising trend of large foundational models in single-cell perturbation modelling inspired by large language models. Lastly, we critically assess the challenges that underline single-cell perturbation modelling while pointing towards relevant future perspectives like perturbation atlases, multi-omics and spatial datasets, causal machine learning for interpretability, multi-task learning for performance and explainability as well as prospects for solving interoperability and benchmarking pitfalls.

Malaria, a significant global health challenge, is caused by Plasmodium parasites. The Plasmodium liver stage plays a pivotal role in the establishment of the infection. This study focuses on the liver stage development of the model organism Plasmodium berghei, employing fluorescent microscopy imaging and convolutional neural networks (CNNs) for analysis. Convolutional neural networks have been recently proposed as a viable option for tasks such as malaria detection, prediction of host-pathogen interactions, or drug discovery. Our research aimed to predict the transition of Plasmodium-infected liver cells to the merozoite stage, a key development phase, 15 hours in advance. We collected and analyzed hourly imaging data over a span of at least 38 hours from 400 sequences, encompassing 502 parasites. Our method was compared to human annotations to validate its efficacy. Performance metrics, including the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity, were evaluated on an independent test dataset. The outcomes revealed an AUC of 0.873, a sensitivity of 84.6%, and a specificity of 83.3%, underscoring the potential of our CNN-based framework to predict liver stage development of P. berghei. These findings not only demonstrate the feasibility of our methodology but also could potentially contribute to the broader understanding of parasite biology.

Accurately predicting molecular properties is a challenging but essential task in drug discovery. Recently, many mono-modal deep learning methods have been successfully applied to molecular property prediction. However, mono-modal learning is inherently limited as it relies solely on a single modality of molecular representation, which restricts a comprehensive understanding of drug molecules. To overcome the limitations, we propose a multimodal fused deep learning (MMFDL) model to leverage information from different molecular representations. Specifically, we construct a triple-modal learning model by employing Transformer-Encoder, Bidirectional Gated Recurrent Unit (BiGRU), and graph convolutional network (GCN) to process three modalities of information from chemical language and molecular graph: SMILES-encoded vectors, ECFP fingerprints, and molecular graphs, respectively. We evaluate the proposed triple-modal model using five fusion approaches on six molecule datasets, including Delaney, Llinas2020, Lipophilicity, SAMPL, BACE, and pKa from DataWarrior. The results show that the MMFDL model achieves the highest Pearson coefficients, and stable distribution of Pearson coefficients in the random splitting test, outperforming mono-modal models in accuracy and reliability. Furthermore, we validate the generalization ability of our model in the prediction of binding constants for protein-ligand complex molecules, and assess the resilience capability against noise. Through analysis of feature distributions in chemical space and the assigned contribution of each modal model, we demonstrate that the MMFDL model shows the ability to acquire complementary information by using proper models and suitable fusion approaches. By leveraging diverse sources of bioinformatics information, multimodal deep learning models hold the potential for successful drug discovery.

Alzheimer’s Disease is the most prevalent neurodegenerative disease, and is a leading cause of disability among the elderly. Eye movement behaviour demonstrates potential as a non-invasive biomarker for Alzheimer’s Disease, with changes detectable at an early stage after initial onset. This paper introduces a new publicly available dataset: EM-COGLOAD (available at https://osf.io/zjtdq/, DOI: 10.17605/OSF.IO/ZJTDQ). A dual-task paradigm was used to create effects of declined cognitive performance in 75 healthy adults as they carried out visual tracking tasks. Their eye movement was recorded, and time series classification of the extracted eye movement traces was explored using a range of deep learning techniques. The results of this showed that convolutional neural networks were able to achieve an accuracy of 87.5% when distinguishing between eye movement under low and high cognitive load, and 76% when distinguishing between the oldest and youngest age groups.

RNA plays an extensive role in a multi-dimensional regulatory system, and its biomedical relationships are scattered across numerous biological studies. However, text mining works dedicated to the extraction of RNA biomedical relations remain limited. In this study, we established a comprehensive and reliable corpus of RNA biomedical relations, recruiting over 30,000 sentences manually curated from more than 15,000 biomedical literature. We also updated RIscoper 2.0, a BERT-based deep learning tool to extract RNA biomedical relation sentences from literature. Benefiting from approximately 100,000 annotated named entities, we integrated the text classification and named entity recognition tasks in this tool. Additionally, RIscoper 2.0 outperformed the original tool in both tasks and can discover new RNA biomedical relations. Additionally, we provided a user-friendly online search tool that enables rapid scanning of RNA biomedical relationships using local and online resources. Both the online tools and data resources of RIscoper 2.0 are available at http://www.rnainter.org/riscoper.

Objectives: Combination therapy of lenvatinib and immune checkpoint inhibitors (CLICI) has emerged as a promising approach for managing unresectable hepatocellular carcinoma (HCC). However, the response to such treatment is observed in only a subset of patients, underscoring the pressing need for reliable methods to identify potential responders. Materials & methods: This was a retrospective analysis involving 120 patients with unresectable HCC. They were divided into training (n = 72) and validation (n = 48) cohorts. We developed an interpretable deep learning model using multiphase computed tomography (CT) images to predict whether patients will respond or not to CLICI treatment, based on the Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST v1.1). We evaluated the models' performance and analyzed the impact of each CT phase. Critical regions influencing predictions were identified and visualized through heatmaps. Results: The multiphase model outperformed the best biphase and uniphase models, achieving an area under the curve (AUC) of 0.802 (95% CI = 0.780–0.824). The portal phase images were found to significantly enhance the model's predictive accuracy. Heatmaps identified six critical features influencing treatment response, offering valuable insights to clinicians. Additionally, we have made this model accessible via a web server at http://uhccnet.com/ for ease of use. Conclusions: The integration of multiphase CT images with deep learning-generated heatmaps for predicting treatment response provides a robust and practical tool for guiding CLICI therapy in patients with unresectable HCC.