CNN-based Model Research Articles

CnnImpute: missing value recovery for single cell RNA sequencing data

The advent of single-cell RNA sequencing (scRNA-seq) technology has revolutionized our ability to explore cellular diversity and unravel the complexities of intricate diseases. However, due to the inherently low signal-to-noise ratio and the presence of an excessive number of missing values, scRNA-seq data analysis encounters unique challenges. Here, we present cnnImpute, a novel convolutional neural network (CNN) based method designed to address the issue of missing data in scRNA-seq. Our approach starts by estimating missing probabilities, followed by constructing a CNN-based model to recover expression values with a high likelihood of being missing. Through comprehensive evaluations, cnnImpute demonstrates its effectiveness in accurately imputing missing values while preserving the integrity of cell clusters in scRNA-seq data analysis. It achieved superior performance in various benchmarking experiments. cnnImpute offers an accurate and scalable method for recovering missing values, providing a useful resource for scRNA-seq data analysis.

Wave Height and Period Estimation from X-Band Marine Radar Images Using Convolutional Neural Network

In this study, a deep learning network for extracting spatial-temporal features is proposed to estimate significant wave height (Hs) and wave period (Ts) from X-band marine radar images. Since the shore-based radar image in this study is interfered with by other radar radial noise lines and solid target objects, to ensure that the proposed convolutional neural network (CNN) extracts the image features accurately, it is necessary to pre-process the radar image to eliminate interference. Firstly, a pre-trained GoogLeNet is used to extract multi-scale depth space features from the radar images to estimate Hs and Ts. Since CNN-based models cannot analyze the temporal behavior of wave features in radar image sequences, self-attention is connected after the deep convolutional layer of the CNN to construct a convolutional self-attention (CNNSA)-based model that generates spatial-temporal features for Hs and Ts estimation. Simultaneously, Hs and Ts measured by nearby buoys are used for model training and reference. The experimental results show that the proposed CNNSA model reduces the RMSD by 0.24 m and 0.11 m, respectively, in Hs estimation compared to the traditional SNR-based and CNN-based methods. In Ts estimation, the RMSD is reduced by 0.3 s and 0.08 s, respectively.

Galaxy stellar and total mass estimation using machine learning

ABSTRACT Conventional galaxy mass estimation methods suffer from model assumptions and degeneracies. Machine learning (ML), which reduces the reliance on such assumptions, can be used to determine how well present-day observations can yield predictions for the distributions of stellar and dark matter. In this work, we use a general sample of galaxies from the TNG100 simulation to investigate the ability of multibranch convolutional neural network (CNN) based ML methods to predict the central (i.e. within 1−2 effective radii) stellar and total masses, and the stellar mass-to-light ratio (M*/L). These models take galaxy images and spatially resolved mean velocity and velocity dispersion maps as inputs. Such CNN-based models can, in general, break the degeneracy between baryonic and dark matter in the sense that the model can make reliable predictions on the individual contributions of each component. For example, with r-band images and two galaxy kinematic maps as inputs, our model predicting M*/L has a prediction uncertainty of 0.04 dex. Moreover, to investigate which (global) features significantly contribute to the correct predictions of the properties above, we utilize a gradient-boosting machine. We find that galaxy luminosity dominates the prediction of all masses in the central regions, with stellar velocity dispersion coming next. We also investigate the main contributing features when predicting stellar and dark matter mass fractions (f*, fDM) and the dark matter mass MDM, and discuss the underlying astrophysics.

Melanoma Skin Cancer Identification with Explainability Utilizing Mask Guided Technique

Melanoma is a highly prevalent and lethal form of skin cancer, which has a significant impact globally. The chances of recovery for melanoma patients substantially improve with early detection. Currently, deep learning (DL) methods are gaining popularity in assisting with the identification of diseases using medical imaging. The paper introduces a computational model for classifying melanoma skin cancer images using convolutional neural networks (CNNs) and vision transformers (ViT) with the HAM10000 dataset. Both approaches utilize mask-guided techniques, employing a specialized U2-Net segmentation module to generate masks. The CNN-based approach utilizes ResNet50, VGG16, and Xception with transfer learning. The training process is enhanced using a Bayesian hyperparameter tuner. Moreover, this study applies gradient-weighted class activation mapping (Grad-CAM) and Grad-CAM++ to generate heatmaps to explain the classification models. These visual heatmaps elucidate the contribution of each input region to the classification outcome. The CNN-based model approach achieved the highest accuracy at 98.37% in the Xception model with a sensitivity and specificity of 95.92% and 99.01%, respectively. The ViT-based model approach achieved high values for accuracy, sensitivity, and specificity, such as 92.79%, 91.09%, and 93.54%, respectively. Furthermore, the performance of the model was assessed through intersection over union (IOU) and other qualitative evaluations. Finally, we developed the proposed model as a web application that can be used as a support tool for medical practitioners in real-time. The system usability study score of 86.87% is reported, which shows the usefulness of the proposed solution.

An Effective Video Transformer With Synchronized Spatiotemporal and Spatial Self-Attention for Action Recognition.

Convolutional neural networks (CNNs) have come to dominate vision-based deep neural network structures in both image and video models over the past decade. However, convolution-free vision Transformers (ViTs) have recently outperformed CNN-based models in image recognition. Despite this progress, building and designing video Transformers have not yet obtained the same attention in research as image-based Transformers. While there have been attempts to build video Transformers by adapting image-based Transformers for video understanding, these Transformers still lack efficiency due to the large gap between CNN-based models and Transformers regarding the number of parameters and the training settings. In this work, we propose three techniques to improve video understanding with video Transformers. First, to derive better spatiotemporal feature representation, we propose a new spatiotemporal attention scheme, termed synchronized spatiotemporal and spatial attention (SSTSA), which derives the spatiotemporal features with temporal and spatial multiheaded self-attention (MSA) modules. It also preserves the best spatial attention by another spatial self-attention module in parallel, thereby resulting in an effective Transformer encoder. Second, a motion spotlighting module is proposed to embed the short-term motion of the consecutive input frames to the regular RGB input, which is then processed with a single-stream video Transformer. Third, a simple intraclass frame interlacing method of the input clips is proposed that serves as an effective video augmentation method. Finally, our proposed techniques have been evaluated and validated with a set of extensive experiments in this study. Our video Transformer outperforms its previous counterparts on two well-known datasets, Kinetics400 and Something-Something-v2.

A convolutional neural network model using intraprostatic patterns of [F18]DCFPyL uptake in PSMA PET images to predict synchronous metastases.

46 Background: [F18]DCFPyL (PyL) is a PSMA targeted imaging agent that provides whole-body staging of prostate cancer. Image analysis of the primary tumor using deep learning algorithms might offer additional insight into disease biology, including co-existing metastatic disease. We developed convolutional neural network (CNN) models using inputs from the entire prostate on PyL images along with auto-segmented hot spots within the of primary prostate cancer tumor to predict synchronous metastases and compared the models against established models built from clinico-pathologic data. Methods: 91 Veterans with de novo prostate cancer were imaged with PyL PSMA PET/CT for initial staging (46% with metastatic disease). The PyL images of the prostate were analyzed using aPROMISE, which autosegments, localizes, and quantifies disease on PSMA PET. The segmentations of the prostate were used to map the PyL PET image of the prostate. Both the entire prostate, as well as aPROMISE defined hot spots were used as inputs for the CNN, where, according to attention map analysis, the hotspot information helps the network understand the location and extent of tumors. The CNN model architecture was based on SqueezeNet v2. The dataset was split into training, validation and test sets using stratified random sampling. The area under ROC curve (AUC) was computed to determine the performance of the model in predicting the presence of metastases and the test predictions were compared with ground truth (M1). Training was repeated 50 times and the best performing experiment was identified. Prediction scores from UCSF-CAPRA and UCLA PSMA risk calculator were used as comparators. Results: The best CNN model had an AUC of 0.89 for prediction of metastatic disease (median 0.72, ICR 0.64 and 0.8). For comparison, the AUC from the UCSF-CAPRA score and UCLA-PSMA risk calculator (any upstaging on PET), which rely on clinicopathologic information, had AUCs of 0.729 and 0.754 in this dataset, respectively. Conclusions: The CNN based model using PyL imaging demonstrates that synchronous metastases can be predicted from intraprostatic PyL uptake patterns alone with a predictive accuracy in this dataset that is at least comparable to published prediction models based on clinicopathologic features. This study raises the hypothesis that PyL CNN-based models could be developed to prognosticate metastatic progression.

Development and prognostic validation of a three-level NHG-like deep learning-based model for histological grading of breast cancer

BackgroundHistological grade is a well-known prognostic factor that is routinely assessed in breast tumours. However, manual assessment of Nottingham Histological Grade (NHG) has high inter-assessor and inter-laboratory variability, causing uncertainty in grade assignments. To address this challenge, we developed and validated a three-level NHG-like deep learning-based histological grade model (predGrade). The primary performance evaluation focuses on prognostic performance.MethodsThis observational study is based on two patient cohorts (SöS-BC-4, N = 2421 (training and internal test); SCAN-B-Lund, N = 1262 (test)) that include routine histological whole-slide images (WSIs) together with patient outcomes. A deep convolutional neural network (CNN) model with an attention mechanism was optimised for the classification of the three-level histological grading (NHG) from haematoxylin and eosin-stained WSIs. The prognostic performance was evaluated by time-to-event analysis of recurrence-free survival and compared to clinical NHG grade assignments in the internal test set as well as in the fully independent external test cohort.ResultsWe observed effect sizes (hazard ratio) for grade 3 versus 1, for the conventional NHG method (HR = 2.60 (1.18–5.70 95%CI, p-value = 0.017)) and the deep learning model (HR = 2.27, 95%CI 1.07–4.82, p-value = 0.033) on the internal test set after adjusting for established clinicopathological risk factors. In the external test set, the unadjusted HR for clinical NHG 2 versus 1 was estimated to be 2.59 (p-value = 0.004) and clinical NHG 3 versus 1 was estimated to be 3.58 (p-value < 0.001). For predGrade, the unadjusted HR for predGrade 2 versus 1 HR = 2.52 (p-value = 0.030), and 4.07 (p-value = 0.001) for preGrade 3 versus 1 was observed in the independent external test set. In multivariable analysis, HR estimates for neither clinical NHG nor predGrade were found to be significant (p-value > 0.05). We tested for differences in HR estimates between NHG and predGrade in the independent test set and found no significant difference between the two classification models (p-value > 0.05), confirming similar prognostic performance between conventional NHG and predGrade.ConclusionRoutine histopathology assessment of NHG has a high degree of inter-assessor variability, motivating the development of model-based decision support to improve reproducibility in histological grading. We found that the proposed model (predGrade) provides a similar prognostic performance as clinical NHG. The results indicate that deep CNN-based models can be applied for breast cancer histological grading.

Deep-learning framework for optimal selection of soil sampling sites

Soil sampling is one of the most fundamental agricultural processes. The selection of soil sampling locations within a field plays a crucial role in soil-health analysis. The current soil-sampling techniques produce random samples across a field, which are not representative of the field conditions. This work leverages the recent advancements of deep learning in image processing for finding optimal locations that present the important characteristics of a field. The data for training are collected at different fields in local farms with five features: aspect, flow accumulation, slope, NDVI (normalized difference vegetation index), and yield. The soil sampling dataset is challenging because of the unbalance between the number of pixels that represent soil-sampling spots and the number of pixels that represent the background area. In this work, we approach the problem with two methods, the first approach involves utilizing a state-of-the-art model with the convolutional neural network (CNN) backbone, while the second is to innovate a deep-learning design grounded in the concepts of transformer and self-attention. Our framework is constructed with an encoder–decoder architecture with the self-attention mechanism as the backbone. In the encoder, the self-attention mechanism is the key feature extractor, which produces feature maps. In the decoder, we introduce atrous convolution networks to concatenate, fuse the extracted features, and then export the optimal locations for soil sampling. Currently, the model has achieved impressive results on the testing dataset, with a mean accuracy of 99.52%, a mean Intersection over Union (IoU) of 57.35%, and a mean Dice Coefficient of 71.47%, while the performance metrics of the state-of-the-art CNN-based model are 66.08%, 3.85%, and 1.98%, respectively. This indicates that our proposed model outperforms the CNN-based method on the soil-sampling dataset. To the best of our knowledge, our work is the first to provide a soil-sampling dataset with multiple attributes and leverage deep learning techniques to enable the automatic selection of soil-sampling sites. This work lays a foundation for novel applications of data science and machine-learning technologies to solve other emerging agricultural problems.

A comparative analysis of CNN-based deep learning architectures for early diagnosis of bone cancer using CT images

Bone cancer is a rare in which cells in the bone grow out of control, resulting in destroying the normal bone tissue. A benign type of bone cancer is harmless and does not spread to other body parts, whereas a malignant type can spread to other body parts and might be harmful. According to Cancer Research UK (2021), the survival rate for patients with bone cancer is 40% and early detection can increase the chances of survival by providing treatment at the initial stages. Prior detection of these lumps or masses can reduce the risk of death and treat bone cancer early. The goal of this current study is to utilize image processing techniques and deep learning-based Convolution neural network (CNN) to classify normal and cancerous bone images. Medical image processing techniques, like pre-processing (e.g., median filter), K-means clustering segmentation, and, canny edge detection were used to detect the cancer region in Computer Tomography (CT) images for parosteal osteosarcoma, enchondroma and osteochondroma types of bone cancer. After segmentation, the normal and cancerous affected images were classified using various existing CNN-based models. The results revealed that AlexNet model showed a better performance with a training accuracy of 98%, validation accuracy of 98%, and testing accuracy of 100%.

Multiclass AI-Generated Deepfake Face Detection Using Patch-Wise Deep Learning Model

In response to the rapid advancements in facial manipulation technologies, particularly facilitated by Generative Adversarial Networks (GANs) and Stable Diffusion-based methods, this paper explores the critical issue of deepfake content creation. The increasing accessibility of these tools necessitates robust detection methods to curb potential misuse. In this context, this paper investigates the potential of Vision Transformers (ViTs) for effective deepfake image detection, leveraging their capacity to extract global features. Objective: The primary goal of this study is to assess the viability of ViTs in detecting multiclass deepfake images compared to traditional Convolutional Neural Network (CNN)-based models. By framing the deepfake problem as a multiclass task, this research introduces a novel approach, considering the challenges posed by Stable Diffusion and StyleGAN2. The objective is to enhance understanding and efficacy in detecting manipulated content within a multiclass context. Novelty: This research distinguishes itself by approaching the deepfake detection problem as a multiclass task, introducing new challenges associated with Stable Diffusion and StyleGAN2. The study pioneers the exploration of ViTs in this domain, emphasizing their potential to extract global features for enhanced detection accuracy. The novelty lies in addressing the evolving landscape of deepfake creation and manipulation. Results and Conclusion: Through extensive experiments, the proposed method exhibits high effectiveness, achieving impressive detection accuracy, precision, and recall, and an F1 rate of 99.90% on a multiclass-prepared dataset. The results underscore the significant potential of ViTs in contributing to a more secure digital landscape by robustly addressing the challenges posed by deepfake content, particularly in the presence of Stable Diffusion and StyleGAN2. The proposed model outperformed when compared with state-of-the-art CNN-based models, i.e., ResNet-50 and VGG-16.

CT synthesis from MR images using frequency attention conditional generative adversarial network

Magnetic resonance (MR) image-guided radiotherapy is widely used in the treatment planning of malignant tumors, and MR-only radiotherapy, a representative of this technique, requires synthetic computed tomography (sCT) images for effective radiotherapy planning. Convolutional neural networks (CNN) have shown remarkable performance in generating sCT images. However, CNN-based models tend to synthesize more low-frequency components and the pixel-wise loss function usually used to optimize the model can result in blurred images. To address these problems, a frequency attention conditional generative adversarial network (FACGAN) is proposed in this paper. Specifically, a frequency cycle generative model (FCGM) is designed to enhance the inter-mapping between MR and CT and extract more rich tissue structure information. Additionally, a residual frequency channel attention (RFCA) module is proposed and incorporated into the generator to enhance its ability in perceiving the high-frequency image features. Finally, high-frequency loss (HFL) and cycle consistency high-frequency loss (CHFL) are added to the objective function to optimize the model training. The effectiveness of the proposed model is validated on pelvic and brain datasets and compared with state-of-the-art deep learning models. The results show that FACGAN produces higher-quality sCT images while retaining clearer and richer high-frequency texture information.

GraphMoCo: A graph momentum contrast model for large-scale binary function representation learning

In the field of cybersecurity, the ability to compute similarity scores at the function level for binary code is of utmost importance. Considering that a single binary file may contain an extensive amount of functions, an effective learning framework must exhibit both high accuracy and efficiency when handling substantial volumes of data. Nonetheless, conventional methods encounter several limitations. Firstly, accurately annotating different pairs of functions with appropriate labels poses a significant challenge, thereby making it difficult to employ supervised learning methods without risk of overtraining. Secondly, while SOTA models often rely on pre-trained encoders or fine-grained graph comparison techniques, these approaches suffer from drawbacks related to time and memory consumption. Thirdly, the momentum update algorithm utilized in graph-based contrastive learning models can result in information leakage. Surprisingly, none of the existing articles address this issue. This research focuses on addressing the challenges associated with large-scale Binary Code Similarity Detection (BCSD). To overcome the aforementioned problems, we propose GraphMoCo: a graph momentum contrast model that leverages multimodal structural information for efficient binary function representation learning on a large scale. We adopt an unsupervised learning strategy. Our approach eliminates the need for manual labeling. By leveraging the intrinsic structural information at multiple levels of the binary code, our model could achieve higher accuracy with a simple CNN-based model. By introducing the preshuffle mechanism, the issue of information leakage in graph momentum update algorithm is mitigated. The evaluation results indicate that GraphMoCo exhibits superior performance compared to SOTA approaches in the function pair search task, showing an average improvement of 7% on AUC, and 10% on MRR and Recall@1. Furthermore, GraphMoCo achieves a MAP of 0.93 on the more challenging dataset 2, which comprises a larger function pool. In a real-world scenario, specifically in known vulnerability searching, GraphMoCo achieves a MRR that surpasses existing SOTA models by 5%.

Use of artificial neural networks in architecture: determining the architectural style of a building with a convolutional neural networks

AbstractThe discussion of "can machines think?" which started with the invention of the modern computer, brought along the question of "can machines design?" by researchers in the design field. These developments in information technologies have also affected the architecture. Artificial intelligence applications are encountered in many areas such as pricing estimation, energy conservation security systems of buildings, ventilation systems, user-oriented interactive design solutions, computer-aided programs used in the plan production phase and design process. When the literature on artificial intelligence applications in the architecture is reviewed, it can be seen that it generally includes shape grammars, graph theory, decision trees, constraint-based models, machine learning methods, RNN (Recursive Neural Networks), CNN (Convolutional Neural Network) and GAN (Generative Adversarial Network) algorithms. In this study, the use of artificial intelligence algorithms in architecture was examined, and an example was designed to determine the architectural structures of different periods by using CNN (Convolutional Neural Network). In the study, the open source TensorFlow library developed by Google and the Python programming language were used. Employing a statistical approach and utilizing convolutional neural networks (CNNs), a study has successfully classified the current flow patterns of buildings based on datasets comprising facades of Gothic, Modern, and Deconstructivist architectural styles. The findings demonstrate the efficacy of CNNs in accurately distinguishing the intricate details of diverse architectural styles. Recognizing elements from different periods using the CNN algorithm can examine not only individual buildings but also the relationship of buildings with their environments. It can also gain an important place in the field of conservation of the architectural discipline. The historical processes, aesthetic features and changes of protected buildings can be learned with the CNN algorithm and can guide restoration decisions. As a result of the study, the employed CNN-based model can correctly classify structures with 84.66% accuracy rate.

Uncovering local aggregated air quality index with smartphone captured images leveraging efficient deep convolutional neural network

The prevalence and mobility of smartphones make these a widely used tool for environmental health research. However, their potential for determining aggregated air quality index (AQI) based on PM2.5 concentration in specific locations remains largely unexplored in the existing literature. In this paper, we thoroughly examine the challenges associated with predicting location-specific PM2.5 concentration using images taken with smartphone cameras. The focus of our study is on Dhaka, the capital of Bangladesh, due to its significant air pollution levels and the large population exposed to it. Our research involves the development of a Deep Convolutional Neural Network (DCNN), which we train using over a thousand outdoor images taken and annotated. These photos are captured at various locations in Dhaka, and their labels are based on PM2.5 concentration data obtained from the local US consulate, calculated using the NowCast algorithm. Through supervised learning, our model establishes a correlation index during training, enhancing its ability to function as a Picture-based Predictor of PM2.5 Concentration (PPPC). This enables the algorithm to calculate an equivalent daily averaged AQI index from a smartphone image. Unlike, popular overly parameterized models, our model shows resource efficiency since it uses fewer parameters. Furthermore, test results indicate that our model outperforms popular models like ViT and INN, as well as popular CNN-based models such as VGG19, ResNet50, and MobileNetV2, in predicting location-specific PM2.5 concentration. Our dataset is the first publicly available collection that includes atmospheric images and corresponding PM2.5 measurements from Dhaka. Our codes and dataset are available at https://github.com/lepotatoguy/aqi.

An attention-based progressive fusion network for pixelwise pavement crack detection

Detecting pavement cracks is a complex task in pavement maintenance. Convolutional neural network (CNN)-based approaches have been increasingly used for this purpose in recent years, but these approaches still have limitations, such as insufficient processing of local features, poor boundary detection ability and information loss issue. In this paper, a novel CNN-based model, named APF-Net, is presented. Specifically, to extract richer and more feature information, proposed model utilizes a progressive fusion module (PF) that enables effective feature enhancement by feature learning from cross-layer features, which operates progressively across adjacent network layers, enabling the crack segmentation model to extract more global and detail features from the crack images. Additionally, a hybrid multiple attention (HMA) module that incorporates both a spatial attention mechanism from two directions and a channel attention mechanism, is proposed, which could effectively capture the long-term dependence of crack features and enhance crack boundary detection. Moreover, proposed model employs a lightweight feature extraction block, named mobile inverted bottleneck convolution (MBC) block, which significantly reduces the model parameters and enables more efficient processing of the crack images. This approach results in a reduction of the computational load and faster detection of pavement cracks. The effectiveness and generalization of the proposed APF-Net have been evaluated through experiments on three publicly available crack datasets. The results demonstrate that the proposed model outperforms other state-of-the-art segmentation models in detecting pavement cracks while with 7.6M parameters and 5.0G FLOPs. Our proposed model can contribute to the development of automatic pavement crack detection systems, which can aid in the maintenance and preservation of roads and highways. The source code for proposed APF-Net is available at https://github.com/MMYZZU/APFNet.git.

A CNN-based model to count the leaves of rosette plants (LC-Net)

Plant image analysis is a significant tool for plant phenotyping. Image analysis has been used to assess plant trails, forecast plant growth, and offer geographical information about images. The area segmentation and counting of the leaf is a major component of plant phenotyping, which can be used to measure the growth of the plant. Therefore, this paper developed a convolutional neural network-based leaf counting model called LC-Net. The original plant image and segmented leaf parts are fed as input because the segmented leaf part provides additional information to the proposed LC-Net. The well-known SegNet model has been utilised to obtain segmented leaf parts because it outperforms four other popular Convolutional Neural Network (CNN) models, namely DeepLab V3+, Fast FCN with Pyramid Scene Parsing (PSP), U-Net, and Refine Net. The proposed LC-Net is compared to the other recent CNN-based leaf counting models over the combined Computer Vision Problems in Plant Phenotyping (CVPPP) and KOMATSUNA datasets. The subjective and numerical evaluations of the experimental results demonstrate the superiority of the LC-Net to other tested models.

A Study on Defect Detection of Dissimilar Joints in Cu-STS Tubes Using Infrared Thermal Imaging of Induction Heating Brazing

We proposed a novel detection method for identifying joint defects in the brazing process between copper tubes and stainless steel using a convolutional neural network (CNN) model. The brazing joints were created using high-frequency induction heating equipment, and infrared thermal imaging cameras were employed to capture the thermal data generated during the jointing process. The experiments involved 15.88 mm diameter copper tubes commonly used in plate heat exchangers, stainless-steel tubes, and filler metal containing 20% Ag. The thermal data were obtained with a resolution of 80 × 80 pixels per frame, resulting in 4796 normal joint data and 5437 defective joint data collected over 100 high-frequency induction-heating brazing experiments. A total of 10,233 thermal imaging data were categorized into 6548 training data, 1638 validation data, and 2047 test data for the development of the predictive model. We designed CNN models with varying hyperparameters, specifically the number of kernel filters and nodes, to evaluate their impact on detection performance. A comparative analysis revealed that a CNN model structure, exhibiting 98.53% accuracy and 99.82% recall on test data, was the most effective. The selected CNN-based defect prediction model demonstrated the potential of using CNN models to discern joint defects in tube configurations that are challenging to identify visually. This study opens avenues for applying CNN-based models for detecting imperfections in complex tube structures.

Multi-label classification of retinal disease via a novel vision transformer model.

The precise identification of retinal disorders is of utmost importance in the prevention of both temporary and permanent visual impairment. Prior research has yielded encouraging results in the classification of retinal images pertaining to a specific retinal condition. In clinical practice, it is not uncommon for a single patient to present with multiple retinal disorders concurrently. Hence, the task of classifying retinal images into multiple labels remains a significant obstacle for existing methodologies, but its successful accomplishment would yield valuable insights into a diverse array of situations simultaneously. This study presents a novel vision transformer architecture called retinal ViT, which incorporates the self-attention mechanism into the field of medical image analysis. To note that this study supposed to prove that the transformer-based models can achieve competitive performance comparing with the CNN-based models, hence the convolutional modules have been eliminated from the proposed model. The suggested model concludes with a multi-label classifier that utilizes a feed-forward network architecture. This classifier consists of two layers and employs a sigmoid activation function. The experimental findings provide evidence of the improved performance exhibited by the suggested model when compared to state-of-the-art approaches such as ResNet, VGG, DenseNet, and MobileNet, on the publicly available dataset ODIR-2019, and the proposed approach has outperformed the state-of-the-art algorithms in terms of Kappa, F1 score, AUC, and AVG.

A deep learning model for brain age prediction using minimally preprocessed T1w images as input.

In the last few years, several models trying to calculate the biological brain age have been proposed based on structural magnetic resonance imaging scans (T1-weighted MRIs, T1w) using multivariate methods and machine learning. We developed and validated a convolutional neural network (CNN)-based biological brain age prediction model that uses one T1w MRI preprocessing step when applying the model to external datasets to simplify implementation and increase accessibility in research settings. Our model only requires rigid image registration to the MNI space, which is an advantage compared to previous methods that require more preprocessing steps, such as feature extraction. We used a multicohort dataset of cognitively healthy individuals (age range = 32.0-95.7 years) comprising 17,296 MRIs for training and evaluation. We compared our model using hold-out (CNN1) and cross-validation (CNN2-4) approaches. To verify generalisability, we used two external datasets with different populations and MRI scan characteristics to evaluate the model. To demonstrate its usability, we included the external dataset's images in the cross-validation training (CNN3). To ensure that our model used only the brain signal on the image, we also predicted brain age using skull-stripped images (CNN4). The trained models achieved a mean absolute error of 2.99, 2.67, 2.67, and 3.08 years for CNN1-4, respectively. The model's performance in the external dataset was in the typical range of mean absolute error (MAE) found in the literature for testing sets. Adding the external dataset to the training set (CNN3), overall, MAE is unaffected, but individual cohort MAE improves (5.63-2.25 years). Salience maps of predictions reveal that periventricular, temporal, and insular regions are the most important for age prediction. We provide indicators for using biological (predicted) brain age as a metric for age correction in neuroimaging studies as an alternative to the traditional chronological age. In conclusion, using different approaches, our CNN-based model showed good performance using one T1w brain MRI preprocessing step. The proposed CNN model is made publicly available for the research community to be easily implemented and used to study ageing and age-related disorders.

Learning discriminative context for salient object detection

Context understanding is important for salient object detection (SOD) in complex scenes. To alleviate visual confusion, we propose a context aware network that full use of the same type contextual information for SOD. Specifically, we introduce the pixel relationships into decoder, which can extract the explicit contextual information to alleviate visual confusion by learning the similarities and differences between pixels. Furthermore, we explore the optimal embedding position of the pixel relationship in network to maximize its benefits, thereby reducing the introduction of noise. Compared with 20 counterparts, experimental results on five datasets show that our approach has better generalization, which averagely improves 1.3% and 2.8% over the advanced CNN-based model and Transformer-based model on three evaluation metrics. The improvements demonstrate effectiveness of the proposed context learning strategies, which are helpful for dealing with various complex scenes. Moreover, our model is more efficient with an inference speed of 56.7 FPS on a single NVIDIA 2080TI GPU. Codes are available at https://github.com/lesonly/CANet.