3D Residual Network Research Articles

Micro-expression recognition based on euler video magnification and 3D residual network under imbalanced sample

A student’s verbal behavior plays a crucial role in education, while nonverbal behavior, such as micro-expressions, significantly improves teaching quality. To address the problem of small facial expression movements, imbalanced data categories, and lack of temporal information in static expressions, a micro-expression recognition method is proposed based on Eulerian Video Magnification (EVM) and a 3D Residual Network (3D ResNet) under imbalanced samples. Firstly, face detection in the Dlib library is used to locate the face in the micro-expression video sample and crop it. Secondly, the EVM is used to magnify the motion features in micro-expressions. Then, the 3D ResNet is used to extract spatio-temporal information from micro-expression video samples, and the Cyclical Focal Loss (CFL) function is introduced in the network training process to solve the class imbalance problem in micro-expression datasets. Finally, the roles of the EVM and the CFL function in recognizing micro-expressions by the 3D ResNet are analyzed. The experimental results on the Spontaneous Micro-expression Database (SMIC) and Chinese Academy of Sciences Micro-expression Database II (CASME II) demonstrate the effectiveness and superiority of this method. The proposed method can assist in teaching evaluation and promote the development of smart classrooms, and further research is needed on the storage and computing of the proposed method on devices.

E-DNet: An End-to-End Dual-Branch Network for Driver Steering Intention Detection

An advanced driving assistant system (ADAS) is critical for improving traffic efficiency and ensuring driving safety. By anticipating the driver’s steering intentions in advance, the system can alert the driver in time to avoid a vehicle collision. This paper proposes a novel end-to-end dual-branch network (EDNet) that utilizes both in-cabin and out-of-cabin data. In this study, we designed an in-cabin driver intent feature extractor based on 3D residual networks and atrous convolution, which is applicable to video data and is capable of capturing a larger range of driver behavior. In order to capture the long-term dependency of temporal data, we designed the depthwise-separable max-pooling (DSMax) module and combined it with a convolutional LSTM to obtain the road environment feature extractor outside the cabin. In addition, to effectively fuse different features inside and outside the cockpit, we designed and propose the dynamic combined-feature attention fusion (D-CAF) module. EDNet employs a freeze-training method, which enables the creation of a lightweight model while simultaneously enhancing the final classification accuracy. Extensive experiments on the Brain4Cars dataset and the Zenodo dataset show that the proposed EDNet was able to recognize the driver’s steering intention up to 3 s in advance. It outperformed the existing state of the art in most driving scenarios.

MMANet: A Multi-Task Residual Network for Alzheimer's Disease Classification and Brain Age Prediction

Objective: Alzheimer's disease (AD) is an irreversible neurodegenerative disease, while mild cognitive impairment (MCI) is a clinical precursor of AD, thus differentiation of AD, MCI and normal control (NC) from noninvasive magnetic resonance imaging (MRI) has positive clinical implications. Material and method: We utilize a 3D residual network to classify AD, MCI, and NC, and add a multiscale module to the original network to enhance the feature representation capability of the network, as well as a cross-dimensional attentional mechanism to enhance the network's attention to important brain regions. We experimentally verified that the network is more inclined to overestimate the brain age of patients in AD and MCI subgroups, thus proving that there is a high correlation between the brain age prediction task and the AD classification task. Therefore, we adopted a multi-task learning approach, using brain age prediction as a supplementary task for AD classification to reduce the risk of overfitting of the network during the training process. Results: Our method achieved 96.02% accuracy, 93.40% precision, 91.48% recall, and 92.24% F1 value in AD/MCI/NC classification. Conclusions: Ablation experiments confirmed that our proposed cross-dimensional attention and multiscale modules can improve the diagnostic performance of AD and MCI, and that multi-task learning in conjunction with brain age prediction can further improve the performance.

Predicting Age from White Matter Diffusivity with Residual Learning.

Imaging findings inconsistent with those expected at specific chronological age ranges may serve as early indicators of neurological disorders and increased mortality risk. Estimation of chronological age, and deviations from expected results, from structural magnetic resonance imaging (MRI) data has become an important proxy task for developing biomarkers that are sensitive to such deviations. Complementary to structural analysis, diffusion tensor imaging (DTI) has proven effective in identifying age-related microstructural changes within the brain white matter, thereby presenting itself as a promising additional modality for brain age prediction. Although early studies have sought to harness DTI's advantages for age estimation, there is no evidence that the success of this prediction is owed to the unique microstructural and diffusivity features that DTI provides, rather than the macrostructural features that are also available in DTI data. Therefore, we seek to develop white-matter-specific age estimation to capture deviations from normal white matter aging. Specifically, we deliberately disregard the macrostructural information when predicting age from DTI scalar images, using two distinct methods. The first method relies on extracting only microstructural features from regions of interest (ROIs). The second applies 3D residual neural networks (ResNets) to learn features directly from the images, which are non-linearly registered and warped to a template to minimize macrostructural variations. When tested on unseen data, the first method yields mean absolute error (MAE) of 6.11 ± 0.19 years for cognitively normal participants and MAE of 6.62 ± 0.30 years for cognitively impaired participants, while the second method achieves MAE of 4.69 ± 0.23 years for cognitively normal participants and MAE of 4.96 ± 0.28 years for cognitively impaired participants. We find that the ResNet model captures subtler, non-macrostructural features for brain age prediction.

Various frameworks for integrating image and video streams for spatiotemporal information learning employing 2D–3D residual networks for human action recognition

Human action recognition has been identified as an important research topic in computer vision because it is an essential form of communication and interplay between computers and humans to assist computers in automatically recognizing human behaviors and accurately comprehending human intentions. Inspired by some keyframe extraction and multifeatured fusion research, this paper improved the accuracy of action recognition by utilizing keyframe features and fusing them with video features. In this article, we suggest a novel multi-stream approach architecture made up of two distinct models fused using different fusion techniques. The first model combines convolutional neural networks in two-dimensional (2D-CNN) with long-short term memory networks to glean long-term spatial and temporal features from video keyframe images for human action recognition. The second model is a three-dimensional convolutional neural network (3D-CNN) that gathers quick spatial–temporal features from video clips. Subsequently, two frameworks are put forth to explain how various fusion structures can improve the performance of action recognition. We investigate methods for video action recognition using early and late fusion. While the late-fusion framework addresses the decision fusion from the two models' choices for action recognition, the early-fusion framework examines the impact of early feature fusion of the two models for action recognition. The various fusion techniques investigate how much each spatial and temporal feature influences the recognition model's accuracy. The HMDB-51 and UCF-101 datasets are two important action recognition benchmarks used to evaluate our method. When applied to the HMDB-51 dataset and the UCF-101 dataset, the early-fusion strategy achieves an accuracy of 70.1 and 95.5%, respectively, while the late-fusion strategy achieves an accuracy of 77.7 and 97.5%, respectively.

Three-Dimensional Residual Neural Architecture Search for Ultrasonic Defect Detection.

This study presents a deep-learning (DL) methodology using 3-D convolutional neural networks (CNNs) to detect defects in carbon fiber-reinforced polymer (CFRP) composites through volumetric ultrasonic testing (UT) data. Acquiring large amounts of ultrasonic training data experimentally is expensive and time-consuming. To address this issue, a synthetic data generation method was extended to incorporate volumetric data. By preserving the complete volumetric data, complex preprocessing is reduced, and the model can utilize spatial and temporal information that is lost during imaging. This enables the model to utilize important features that might be overlooked otherwise. The performance of three architectures was compared. The first architecture is prevalent in the literature for the classification of volumetric datasets. The second demonstrated a hand-designed approach to architecture design, with modifications to the first architecture to address the challenges of this specific task. A key modification was the use of cuboidal kernels to account for the large aspect ratios seen in ultrasonic data. The third architecture was discovered through neural architecture search (NAS) from a modified 3-D residual neural network (ResNet) search space. In addition, domain-specific augmentation methods were incorporated during training, resulting in significant improvements in model performance, with a mean accuracy improvement of 22.4% on the discovered architecture. The discovered architecture demonstrated the best performance with a mean accuracy increase of 7.9% over the second-best model. It was able to consistently detect all defects while maintaining a model size smaller than most 2-D ResNets. Each model had an inference time of less than 0.5 s, making them efficient for the interpretation of large amounts of data.

AST3DRNet: Attention-Based Spatio-Temporal 3D Residual Neural Networks for Traffic Congestion Prediction.

Traffic congestion prediction has become an indispensable component of an intelligent transport system. However, one limitation of the existing methods is that they treat the effects of spatio-temporal correlations on traffic prediction as invariable during modeling spatio-temporal features, which results in inadequate modeling. In this paper, we propose an attention-based spatio-temporal 3D residual neural network, named AST3DRNet, to directly forecast the congestion levels of road networks in a city. AST3DRNet combines a 3D residual network and a self-attention mechanism together to efficiently model the spatial and temporal information of traffic congestion data. Specifically, by stacking 3D residual units and 3D convolution, we proposed a 3D convolution module that can simultaneously capture various spatio-temporal correlations. Furthermore, a novel spatio-temporal attention module is proposed to explicitly model the different contributions of spatio-temporal correlations in both spatial and temporal dimensions through the self-attention mechanism. Extensive experiments are conducted on a real-world traffic congestion dataset in Kunming, and the results demonstrate that AST3DRNet outperforms the baselines in short-term (5/10/15 min) traffic congestion predictions with an average accuracy improvement of 59.05%, 64.69%, and 48.22%, respectively.

Image-based 3D reconstruction and permeability modelling of rock using enhanced interpretable deep residual learning

The study presents a novel deep residual learning framework with interpretability for the prediction of rock permeability. The proposed framework, termed ResNet-NiN integrates a 3D deep residual network (ResNet) with network in network (NiN) connections. A total of 1331 datasets were generated using voxel segmentation of DRP samples obtained from actual Berea Sandstone and subsequently, a percolation simulation was performed. The learning rate decay strategy of Cosine Annealing with warmup restarts is used to train the predictive model and prevent the predicted permeability from being trapped in the local optima. To improve the trust and comprehensibility of the model and visualize the ResNet-NiN architecture, we established a compound approach that leverages the internal gradient change of the neural network into the deep learning framework. This integrated approach encompasses the analysis of Integrated Gradients (IG) from the input perspective and Gradient-weighted Class Activation Mapping (Grad-CAM) from the output perspective. The experimental findings demonstrate that the developed model exhibits superiority over the typically employed 3D convolutional neural network architecture in terms of both accuracy and stability. The unified interpretability technique offers a coherent explanation for the provided sample, validating the model's capability and elucidating the factors and relationships underlying the model's prediction.

A CT-based deep learning model: visceral pleural invasion and survival prediction in clinical stage IA lung adenocarcinoma

Pathologic visceral pleural invasion (VPI) in patients with early-stage lung cancer can result in the upstaging of T1 to T2, in addition to having implications for surgical resection and prognostic outcomes. This study was designed with the goal of establishing and validating a CT-based deep learning (DL) model capable of predicting VPI status and stratifying patients based on their prognostic outcomes. In total, 2077 patients from three centers with pathologically confirmed clinical stage IA lung adenocarcinoma were enrolled. DL signatures were extracted with a 3D residual neural network. DL model was able to effectively predict VPI status. VPI predicted by the DL models, as well as pathologic VPI, was associated with shorter disease-free survival. The established deep learning signature provides a tool capable of aiding the accurate prediction of VPI in patients with clinical stage IA lung adenocarcinoma, thus enabling prognostic stratification.

Integration of dosimetric parameters, clinical factors, and radiomics to predict symptomatic radiation pneumonitis in lung cancer patients undergoing combined immunotherapy and radiotherapy

PurposeThis study aimed to combine clinical/dosimetric factors and handcrafted/deep learning radiomic features to establish a predictive model for symptomatic (grade ≥ 2) radiation pneumonitis (RP) in lung cancer patients who received immunotherapy followed by radiotherapy. Materials and MethodsThis study retrospectively collected data of 73 lung cancer patients with prior receipt of ICIs who underwent thoracic radiotherapy (TRT). Of these 73 patients, 41 (56.2 %) developed symptomatic grade ≥ 2 RP. RP was defined per multidisciplinary clinician consensus using CTCAE v5.0. Regions of interest (ROIs) (from radiotherapy planning CT images) utilized herein were gross tumor volume (GTV), planning tumor volume (PTV), and PTV-GTV. Clinical/dosimetric (mean lung dose and V5-V30) parameters were collected, and 107 handcrafted radiomic (HCR) features were extracted from each ROI. Deep learning-based radiomic (DLR) features were also extracted based on pre-trained 3D residual network models. HCR models, Fusion HCR model, Fusion HCR + ResNet models, and Fusion HCR + ResNet + Clinical models were built and compared using the receiver operating characteristic (ROC) curve with measurement of the area under the curve (AUC). Five-fold cross-validation was performed to avoid model overfitting. ResultsHCR models across various ROIs and the Fusion HCR model showed good predictive ability with AUCs from 0.740 to 0.808 and 0.740–0.802 in the training and testing cohorts, respectively. The addition of DLR features improved the effectiveness of HCR models (AUCs from 0.826 to 0.898 and 0.821–0.898 in both respective cohorts). The best performing prediction model (HCR + ResNet + Clinical) combined HCR & DLR features with 7 clinical/dosimetric characteristics and achieved an average AUC of 0.936 and 0.946 in both respective cohorts. ConclusionsIn patients undergoing combined immunotherapy/RT for lung cancer, integrating clinical/dosimetric factors and handcrafted/deep learning radiomic features can offer a high predictive capacity for RP, and merits further prospective validation.

EXTRACTION OF PAINT LOSS IN ANCIENT MURALS BASED ON 3D RESIDUAL NEURAL NETWORK

Abstract. The ancient murals of Qutan Temple in Qinghai Province have a very serious loss of paint. Moreover, the main components of the base color paint layer in the paint loss area and the white patterns in the murals are both calcified, which are similar in color and spectral features. Thus, it is difficult to distinguish them by only using spectral features. A method of paint loss area extraction based on 3D residual network with multi-scale feature fusion is proposed. Firstly, the hyperspectral images with paint loss regions were collected by hyperspectral images. They are pre-processed to establish the training data set. Secondly, 3D residual network models are constructed using 3×3×3, 3×3×5 and 5×5×3 convolution kernels to realize the extraction and fusion of spatial and spectral features at different scales of hyperspectral images. The produced mural hyperspectral dataset is used for network training to obtain the prediction model. Finally, the hyperspectral images are input into the trained model to achieve the extraction of paint loss. After comparing different methods, the experimental result shows that the proposed method can improve the extraction accuracy of mural paint loss and serve as a reference for other deteriorations extraction.

Abstract 352: Automated Deep Learning Algorithms Outperforms Gray-White Matter Ratio for the Detection of Severe Cerebral Edema in Cardiac Arrest Survivors

Introduction: Significant knowledge gaps remain in predicting neurological outcomes for comatose cardiac arrest survivors. Severe cerebral edema (SCE) development offers an objective imaging surrogate for neurologic injury. Aim: We have previously shown that deep learning (DL) algorithms can detect SCE on CT. We aim to determine whether our model will perform equivalently on an independent external cohort. We also compared the performance of the DL model with one based on the gray-white matter ratio (GWR) for detecting SCE. Method: We retrospectively analyzed CT data (299 patients, 55±17 years, 71% male, 65 with SCE). Scans whose radiology reports described herniation or more than minimal ventricular effacement were classified as SCE. Datasets were randomly split (balancing for SCE) into training, validation, and testing cohorts. We used transfer learning with a 3D residual neural network framework to train our model. The probability of SCE was calculated. We applied the model to an external cohort (151 patients, 58±17 years, 63% male, 29 with SCE). Regions of interest (ROIs) were drawn in the caudate nuclei (CN) and posterior limbs of the internal capsule (PLIC). GWR was calculated as the ratio of mean x-ray attenuation values in the CN and the PLIC. Areas under the receiver operating curves (AUC) for the DL and GWR models were compared (bootstrap test). Results: The optimal model achieved comparable accuracies for the internal validation cohort (84% [95% CI: 78-89%], n=60) and reserved test cohort (83% [95% CI: 76 to 89%], n=61). The external cohort consisted of 175 scans. The model’s accuracy was 87% [95% CI: 81% to 91%], comparable to the results from the testing and internal validation cohorts. AUC for DL was significantly greater than for GWR (92% vs. 73%, P=0.001) (see Figure). Conclusions: Our result shows that deep transfer learning is a promising method to automate SCE detection on CT images and outperforms GWR measurements.

Diffusion Convolution Neural Network-Based Multiview Gesture Recognition for Athletes in Dynamic Scenes

This paper focuses on deep vision sensing-assisted gesture recognition for athletes in dynamic scenes. Although many research attention had been devoted to this field in recent years, most of existing works failed to fully take characteristics of dynamic scenes into consideration. To deal with this challenge, this paper proposes a diffusion convolution neural network-based multiview gesture recognition approach in dynamic scenes. For one thing, the dynamic spatiotemporal slice position selection based on the body mask heatmap is adopted to calculate positions of horizontal and vertical slices. Thus, the dynamic selection of slice positions in two directions can be realized, and then the extraction of bi-directional spatiotemporal slice images can be completed. For another, action sequences through the 3D residual neural network are learned, and the spatiotemporal information among frames are mined through recurrent networks. Through their combination, a multi-view gesture recognition approach for athletes is constructed. In the experiments, two standard datasets UCF101 and HMDB51 are utilized to establish simulation environment. The proposed method can reach the accuracy beyond 95% on the two datasets. Compared with several typical recognition methods, the proposed method shows higher accuracy.

Airborne hyperspectral imaging for early diagnosis of kimchi cabbage downy mildew using 3D-ResNet and leaf segmentation

Kimchi cabbage (Brassica rapa pekinensis), one of the main agricultural products in Korea, is susceptible to downy mildew disease infections. Infected plants develop yellow spots (chlorosis) on the upper (adaxial) side of the infected leaf, undermining cabbage production and quality. An early detection method to recognize and treat the disease is crucial to prevent downy mildew and lessen its physical effects on plants. Hyperspectral imaging can capture data from a broad spectrum, which can be utilized to detect disease occurrence before any visible symptoms appear. Combining a hyperspectral camera with an unmanned aerial vehicle (UAV) can provide a non-destructive, field-scale disease detection system. In this study, three-dimensional (3D) convolutional neural network (CNN) models were used to simultaneously account for the spectral and spatial features of the disease to enable automatic disease detection. Using a 3D-residual network (ResNet) CNN with four residual blocks, each followed by a rectified linear unit activation function and a max-pooling layer, helped achieve an overall accuracy of 0.876 and a diseased class accuracy of 0.873. Disease severity was estimated by grouping nearby diseased leaves using the density-based spatial clustering of applications with noise clustering algorithm to achieve a 27.07 % relative error or a 1.08 level difference from the actual.

Three-dimensional ultrasound image reconstruction based on 3D-ResNet in the musculoskeletal system using a 1D probe: ex vivo and in vivo feasibility studies

Objective. Three-dimensional (3D) ultrasound (US) is needed to provide sonographers with a more intuitive panoramic view of the complex anatomical structure, especially the musculoskeletal system. In actual scanning, sonographers may perform fast scanning using a one-dimensional (1D) array probe .at random angles to gain rapid feedback, which leads to a large US image interval and missing regions in the reconstructed volume. Approach. In this study, a 3D residual network (3D-ResNet) modified by a 3D global residual branch (3D-GRB) and two 3D local residual branches (3D-LRBs) was proposed to retain detail and reconstruct high-quality 3D US volumes with high efficiency using only sparse two-dimensional (2D) US images. The feasibility and performance of the proposed algorithm were evaluated on ex vivo and in vivo sets. Main r esults. High-quality 3D US volumes in the fingers, radial and ulnar bones, and metacarpophalangeal joints were obtained by the 3D-ResNet, respectively. Their axial, coronal, and sagittal slices exhibited rich texture and speckle details. Compared with kernel regression, voxel nearest-neighborhood, squared distance weighted methods, and a 3D convolution neural network in the ablation study, the mean peak-signal-to-noise ratio and mean structure similarity of the 3D-ResNet were up to 28.53 ± 1.29 dB and 0.98 ± 0.01, respectively, and the corresponding mean absolute error dropped to 0.023 ± 0.003 with a better resolution gain of 1.22 ± 0.19 and shorter reconstruction time. Significance. These results illustrate that the proposed algorithm can rapidly reconstruct high-quality 3D US volumes in the musculoskeletal system in cases of a large amount of data loss. This suggests that the proposed algorithm has the potential to provide rapid feedback and precise analysis of stereoscopic details in complex and meticulous musculoskeletal system scanning with a less limited scanning speed and pose variations for the 1D array probe.

Spatial-spectral attention-enhanced Res-3D-OctConv for corn and weed identification utilizing hyperspectral imaging and deep learning

Corn production is an important basis to ensure the world food security, and weeds in the field will cause corn production decline. Therefore, in order to quickly recognize corn and weed in the field, a model was proposed by combining hyperspectral image with deep learning method. However, there are some problems in hyperspectral image, such as high redundancy of adjacent spectra and insufficient feature information extraction. In order to solve the above problems, the four principal components based on principal component analysis (PCA) were firstly extracted in this paper, so as to decrease the information redundancy between adjacent spectra. Secondly, the residual three-dimensional octave convolution (Res-3D-OctConv) was used to excavate the spatial information from the frequency components, while taking into account the spectral information. Finally, spatial and spectral attention models were introduced to highlight important spatial information and spectral information. At the same time, the spatial information and spectral information was integrated by cross fusion. Experimental results show that the recognition accuracy of the proposed model is 98.56 %, which is 8.65 % and 10.20 % higher than that of k-nearest neighbor (KNN) and support vector machine (SVM) respectively. The recognition result of the proposed model is further compared with that of 3D residual network (3D-ResNet) and 3D convolutional neural network (3D-CNN), and the recognition precision of the proposed model in this paper is increased by 1.40 % and 1.02 % compared with 3D-CNN and 3D-ResNet, respectively. The results show that the proposed model can better recognize the hyperspectral images of corn and weed.

PolSAR Image Classification Based on Relation Network with SWANet

Deep learning and convolutional neural networks (CNN) have been widely applied in polarimetric synthetic aperture radar (PolSAR) image classification, and satisfactory results have been obtained. However, there is one crucial issue that still has not been solved. These methods require abundant labeled samples and obtaining the labeled samples of PolSAR images is usually time-consuming and labor-intensive. To obtain better classification results with fewer labeled samples, a new attention-based 3D residual relation network (3D-ARRN) is proposed for PolSAR image. Firstly, a multilayer CNN with residual structure is used to extract depth polarimetric features. Secondly, to extract more important feature information and improve the classification results, a spatial weighted attention network (SWANet) is introduced to concentrate the feature information, which is more favorable for a classification task. Then, the features of training and test samples are integrated and CNN is utilized to compute the score of similarity between training and test samples. Finally, the similarity score is used to determine the category of test samples. Studies on four different PolSAR datasets illustrate that the proposed 3D-ARRN model can achieve higher classification results than other comparison methods with few labeled data.

Prediction of lymph node metastasis in stage T1–2 rectal cancers with MRI-based deep learning

This study aimed to investigate whether a deep learning (DL) model based on preoperative MR images of primary tumors can predict lymph node metastasis (LNM) in patients with stage T1-2 rectal cancer. In this retrospective study, patients with stage T1-2 rectal cancer who underwent preoperative MRI between October 2013 and March 2021 were included and assigned to the training, validation, and test sets. Four two-dimensional and three-dimensional (3D) residual networks (ResNet18, ResNet50, ResNet101, and ResNet152) were trained and tested on T2-weighted images to identify patients with LNM. Three radiologists independently assessed LN status on MRI, and diagnostic outcomes were compared with the DL model. Predictive performance was assessed with AUC and compared using the Delong method. In total, 611 patients were evaluated (444 training, 81 validation, and 86 test). The AUCs of the eight DL models ranged from 0.80 (95% confidence interval [CI]: 0.75, 0.85) to 0.89 (95% CI: 0.85, 0.92) in the training set and from 0.77 (95% CI: 0.62, 0.92) to 0.89 (95% CI: 0.76, 1.00) in the validation set. The ResNet101 model based on 3D network architecture achieved the best performance in predicting LNM in the test set, with an AUC of 0.79 (95% CI: 0.70, 0.89) that was significantly greater than that of the pooled readers (AUC, 0.54 [95% CI: 0.48, 0.60]; p < 0.001). The DL model based on preoperative MR images of primary tumors outperformed radiologists in predicting LNM in patients with stage T1-2 rectal cancer. • Deep learning (DL) models with different network frameworks showed different diagnostic performance for predicting lymph node metastasis (LNM) in patients with stage T1-2 rectal cancer. • The ResNet101 model based on 3D network architecture achieved the best performance in predicting LNM in the test set. • The DL model based on preoperative MR images outperformed radiologists in predicting LNM in patients with stage T1-2 rectal cancer.

The application of video text generation technology in assessing the effectiveness of teaching ethnic traditional sports

Abstract Ethnic traditional sports are forms of physical activity unique to one or more ethnic groups. Effective evaluation of the teaching effectiveness of ethnic traditional sports helps to promote the development of ethnic sports and the transmission of ethnic traditional sports. Currently, the evaluation of ethnic traditional sports is mainly a scoring system, which evaluates students’ performance in ethnic traditional sports through teachers’ scoring, and this evaluation method is difficult to assess the teaching effectiveness of ethnic traditional sports accurately. In this paper, based on the video text generation technology, the 3D residual attention network model is constructed by introducing the attention mechanism into the 3D residual module in the video feature extraction through the encoding-decoding video text description framework and improving the traditional deep residual network for evaluating the teaching effect of ethnic sports. After experimental validation, it is shown that the 3D residual attention network model can more accurately describe the evaluation of ethnic traditional sports teaching effectiveness using natural language.

A comprehensive characterization of hippocampal feature ensemble serves as individualized brain signature for Alzheimer's disease: deep learning analysis in 3238 participants worldwide.

Hippocampal characterization is one of the most significant hallmarks of Alzheimer's disease (AD); rather, the single-level feature is insufficient. A comprehensive hippocampal characterization is pivotal for developing a well-performing biomarker for AD. To verify whether a comprehensive characterization of hippocampal features of gray matter volume, segmentation probability, and radiomics features could better distinguish AD from normal control (NC), and to investigate whether the classification decision score could serve as a robust and individualized brain signature. A total of 3238 participants' structural MRI from four independent databases were employed to conduct a 3D residual attention network (3DRA-Net) to classify NC, mild cognitive impairment (MCI), and AD. The generalization was validated under inter-database cross-validation. The neurobiological basis of the classification decision score as a neuroimaging biomarker was systematically investigated by association with clinical profiles, as well as longitudinal trajectory analysis to reveal AD progression. All image analyses were performed only upon the single modality of T1-weighted MRI. Our study exhibited an outstanding performance (ACC = 91.6%, AUC = 0.95) of the comprehensive characterization of hippocampal features in distinguishing AD (n = 282) from NC (n = 603) in Alzheimer's Disease Neuroimaging Initiative cohort, and ACC = 89.2% and AUC = 0.93 under external validation. More importantly, the constructed score was significantly correlated with clinical profiles (p < 0.05), and dynamically altered over the AD longitudinal progression, provided compelling evidence of a solid neurobiological basis. This systemic study highlights the potential of the comprehensive characterization of hippocampal features to provide an individualized, generalizable, and biologically plausible neuroimaging biomarker for early detection of AD. • The comprehensive characterization of hippocampal features exhibited ACC = 91.6% (AUC = 0.95) in classifying AD from NC under intra-database cross-validation, and ACC = 89.2% (AUC = 0.93) in external validation. • The constructed classification score was significantly associated with clinical profiles, and dynamically altered over the AD longitudinal progression, which highlighted its potential of being an individualized, generalizable, and biologically plausible neuroimaging biomarker for early detection of AD.