Deep Convolutional Neural Network Model Research Articles

Deep Learning Based Polycystic Ovary Syndrome (PCOS) Identification

Polycystic ovary syndrome (PCOS) is a common endocrine disorder affecting reproductive-aged women, characterized by hormonal imbalance and ovarian cysts. Timely and accurate diagnosis of PCOS is crucial for effective management and prevention of associated complications. In this paper, we propose an automated diagnostic system for PCOS using deep learning and convolutional neural networks (CNNs). The dataset comprises ultrasound images of ovaries categorized into infected (PCOS) and not infected (non-PCOS) classes. Two CNN architectures, a deep CNN model and a CNN model with a last layer replaced by a support vector machine (SVM), are implemented and evaluated. Evaluation metrics include accuracy, precision, recall, and F1-score, demonstrating promising performance in accurately distinguishing between PCOS and non-PCOS cases. This automated diagnostic system offers a reliable and efficient approach for PCOS diagnosis, potentially reducing dependence on manual interpretation of ultrasound scans and improving healthcare outcomes for affected individuals. Key Words: Polycystic ovary syndrome (PCOS) , Deep learning, Convolutional neural networks (CNNs), Data augmentation, Ultrasound Scans

Preliminary study on AI-assisted diagnosis of bone remodeling in chronic maxillary sinusitis

ObjectiveTo construct the deep learning convolution neural network (CNN) model and machine learning support vector machine (SVM) model of bone remodeling of chronic maxillary sinusitis (CMS) based on CT image data to improve the accuracy of image diagnosis.MethodsMaxillary sinus CT data of 1000 samples in 500 patients from January 2018 to December 2021 in our hospital was collected. The first part is the establishment and testing of chronic maxillary sinusitis detection model by 461 images. The second part is the establishment and testing of the detection model of chronic maxillary sinusitis with bone remodeling by 802 images. The sensitivity, specificity and accuracy and area under the curve (AUC) value of the test set were recorded, respectively.ResultsPreliminary application results of CT based AI in the diagnosis of chronic maxillary sinusitis and bone remodeling. The sensitivity, specificity and accuracy of the test set of 93 samples of CMS, were 0.9796, 0.8636 and 0.9247, respectively. Simultaneously, the value of AUC was 0.94. And the sensitivity, specificity and accuracy of the test set of 161 samples of CMS with bone remodeling were 0.7353, 0.9685 and 0.9193, respectively. Simultaneously, the value of AUC was 0.89.ConclusionIt is feasible to use artificial intelligence research methods such as deep learning and machine learning to automatically identify CMS and bone remodeling in MSCT images of paranasal sinuses, which is helpful to standardize imaging diagnosis and meet the needs of clinical application.

MEGANET: Improved framework with nature inspired approach for colorectal cancer polyp classification

BACKGROUND: Polyps are tumorous growths in the colon or rectal area which can turn into cancer at later stages, thus detection of polyps is very important for timely prevention of colorectal cancer. The aim of the study is to develop a framework to accurately detect polyp images in colonoscopy images. OBJECTIVE: Development of an intelligent framework for classification of colorectal cancer from colon and rectal images. The standard machine learning, convolutional neural networks and ensemble models with nature inspired approach were implemented for this study. Model optimization was performed by varying hyper parameters. The main objective was to find an optimal model with high accuracy, optimized weights and less parameters. METHODS: The deep learning Convolutional Neural Network (CNN) models such as VGG19, ResNet50, EfficientNet, Ensemble Model (EM), and Modified Ensemble CNN with Genetic Algorithm (MEGANET) were implemented for the classification of colon images. RESULTS: Ensemble model was also created with two best performing deep learning models to further achieve higher accuracy of 96%. The ensemble model outperformed the other models in terms of accuracy, precision, recall, and F1 score. But this model has more complexity. The MEGANET, nature inspired evolutionary ensemble CNN model was implemented with transfer learning and genetic algorithms for weights optimization and parameter reduction. It achieved accuracy of 95%, on training data. CONCLUSION: The MEGANET performed similar to EM with less number of parameters on training, validation and test dataset. In future different methods will be implemented to further reduce the parameters and attain reasonable accuracy using MEGANET.

Enhancing groundwater level prediction accuracy using interpolation techniques in deep learning models

Groundwater surface (GWS), which denotes the vertical extent of the water table or the volume of subterranean water within geologic formations, is pivotal for effective groundwater resource management. Accurately predicting GWS requires comprehensive and precise data to fully understand the influencing factors. The inherent temporal complexity and often incomplete datasets of GWS pose significant challenges to accurate assessments. This research aims to devise a comprehensive method that merges interpolation and prediction techniques to develop a functional model and dynamic system for GWS prediction. The study was conducted on the Azarshahr Plain aquifer in Iran, involving 34 observation wells with partially or entirely missing data. Initial analysis utilized three interpolation methods—Kriging, Support Vector Machine (SVM), and M5P—with the M5P method emerging as the most accurate, evidenced by the lowest Root Mean Square Error (RMSE) of 1.83. Two subsequent scenarios were examined: (1) using the M5P method to interpolate missing data for all 34 wells, and (2) using only data from 15 wells with complete records. GWS levels were predicted using Deep Neural Network (DNN) and Convolutional Neural Network (CNN) models. Comparative analysis highlighted the superior performance of the CNN model in both scenarios, particularly noting its effectiveness in GWS prediction. The improvement of data quality through interpolation significantly enhanced predictive accuracy by approximately 90 percent, thereby increasing the reliability of the predictive models for future groundwater management decisions.

Optimizing motor imagery BCI models with hard trials removal and model refinement

Deep learning models have demonstrated remarkable performance in the classification of motor imagery BCI systems. However, these models exhibit sensitivity to challenging trials, often called hard trials, leading to performance degradation. In this paper, we address this issue by proposing two novel methods for identifying and mitigating the impact of hard trials on model performance. The first method leverages model prediction scores to discern hard trials. The second approach employs a quantitative explainable artificial intelligence (XAI) approach, enabling a more transparent and interpretable means of hard trial identification. The identified hard trials are removed from the entire motor imagery training and validation dataset, and the deep learning model is further re-trained using the dataset without hard trials. To evaluate the efficacy of these proposed methods, experiments were conducted on the Open BMI dataset. The results for hold-out analysis show that the proposed quantitative XAI- based hard trial removal method has statistically improved the average classification accuracy of the baseline deep CNN model from 63.77% to 68.70%, with p-value = 7.66e−11 for the subject-specific MI classification. Additionally, analyzing the scalp map representing the average relevance scores of correctly classified trials compared to a misclassified trial provides a deeper insight into identifying hard trials. The results indicate that the proposed quantitaive-based XAI approach outperformes the prediction-score-based approach in hard trial identification.

Deep learning-based risk stratification of preoperative breast biopsies using digital whole slide images

BackgroundNottingham histological grade (NHG) is a well established prognostic factor in breast cancer histopathology but has a high inter-assessor variability with many tumours being classified as intermediate grade, NHG2. Here, we evaluate if DeepGrade, a previously developed model for risk stratification of resected tumour specimens, could be applied to risk-stratify tumour biopsy specimens.MethodsA total of 11,955,755 tiles from 1169 whole slide images of preoperative biopsies from 896 patients diagnosed with breast cancer in Stockholm, Sweden, were included. DeepGrade, a deep convolutional neural network model, was applied for the prediction of low- and high-risk tumours. It was evaluated against clinically assigned grades NHG1 and NHG3 on the biopsy specimen but also against the grades assigned to the corresponding resection specimen using area under the operating curve (AUC). The prognostic value of the DeepGrade model in the biopsy setting was evaluated using time-to-event analysis.ResultsBased on preoperative biopsy images, the DeepGrade model predicted resected tumour cases of clinical grades NHG1 and NHG3 with an AUC of 0.908 (95% CI: 0.88; 0.93). Furthermore, out of the 432 resected clinically-assigned NHG2 tumours, 281 (65%) were classified as DeepGrade-low and 151 (35%) as DeepGrade-high. Using a multivariable Cox proportional hazards model the hazard ratio between DeepGrade low- and high-risk groups was estimated as 2.01 (95% CI: 1.06; 3.79).ConclusionsDeepGrade provided prediction of tumour grades NHG1 and NHG3 on the resection specimen using only the biopsy specimen. The results demonstrate that the DeepGrade model can provide decision support to identify high-risk tumours based on preoperative biopsies, thus improving early treatment decisions.

Traffic Sign Recognition using HIICNN Stack Ensemble Method

Abstract: This project introduces a novel deep convolutional neural network (CNN) model aimed at enhancing traffic sign recognition for autonomous vehicle technology and road safety. Employing a stacking ensemble approach, we combine multiple CNNs with diverse improvement techniques, including preprocessing for adverse weather conditions and shooting errors, and data augmentation to handle class imbalance. Through adjustments in learning rates during model training, we mitigate overfitting. Our ensemble model achieves an exceptional 99.4 percent test accuracy on the German Traffic Sign Recognition Benchmark (GTSRB) dataset, surpassing prior studies. We utilize gradient-weighted Class Activation Mapping (Grad-CAM) for model explain- ability and an evidential deep learning approach to quantify classification uncertainty. Our framework underscores the efficacy of combining preprocessing, ensemble learning, and transfer learning for superior perfor- mance and reliability in traffic sign recognition systems, contributing significantly to autonomous driving and road safety.

A Comparative Analysis of the Novel Conditional Deep Convolutional Neural Network Model, Using Conditional Deep Convolutional Generative Adversarial Network-Generated Synthetic and Augmented Brain Tumor Datasets for Image Classification.

Disease prediction is greatly challenged by the scarcity of datasets and privacy concerns associated with real medical data. An approach that stands out to circumvent this hurdle is the use of synthetic data generated using Generative Adversarial Networks (GANs). GANs can increase data volume while generating synthetic datasets that have no direct link to personal information. This study pioneers the use of GANs to create synthetic datasets and datasets augmented using traditional augmentation techniques for our binary classification task. The primary aim of this research was to evaluate the performance of our novel Conditional Deep Convolutional Neural Network (C-DCNN) model in classifying brain tumors by leveraging these augmented and synthetic datasets. We utilized advanced GAN models, including Conditional Deep Convolutional Generative Adversarial Network (DCGAN), to produce synthetic data that retained essential characteristics of the original datasets while ensuring privacy protection. Our C-DCNN model was trained on both augmented and synthetic datasets, and its performance was benchmarked against state-of-the-art models such as ResNet50, VGG16, VGG19, and InceptionV3. The evaluation metrics demonstrated that our C-DCNN model achieved accuracy, precision, recall, and F1 scores of 99% on both synthetic and augmented images, outperforming the comparative models. The findings of this study highlight the potential of using GAN-generated synthetic data in enhancing the training of machine learning models for medical image classification, particularly in scenarios with limited data available. This approach not only improves model accuracy but also addresses privacy concerns, making it a viable solution for real-world clinical applications in disease prediction and diagnosis.

Enhancing oral squamous cell carcinoma detection: a novel approach using improved EfficientNet architecture

ProblemOral squamous cell carcinoma (OSCC) is the eighth most prevalent cancer globally, leading to the loss of structural integrity within the oral cavity layers and membranes. Despite its high prevalence, early diagnosis is crucial for effective treatment.AimThis study aimed to utilize recent advancements in deep learning for medical image classification to automate the early diagnosis of oral histopathology images, thereby facilitating prompt and accurate detection of oral cancer.MethodsA deep learning convolutional neural network (CNN) model categorizes benign and malignant oral biopsy histopathological images. By leveraging 17 pretrained DL-CNN models, a two-step statistical analysis identified the pretrained EfficientNetB0 model as the most superior. Further enhancement of EfficientNetB0 was achieved by incorporating a dual attention network (DAN) into the model architecture.ResultsThe improved EfficientNetB0 model demonstrated impressive performance metrics, including an accuracy of 91.1%, sensitivity of 92.2%, specificity of 91.0%, precision of 91.3%, false-positive rate (FPR) of 1.12%, F1 score of 92.3%, Matthews correlation coefficient (MCC) of 90.1%, kappa of 88.8%, and computational time of 66.41%. Notably, this model surpasses the performance of state-of-the-art approaches in the field.ConclusionIntegrating deep learning techniques, specifically the enhanced EfficientNetB0 model with DAN, shows promising results for the automated early diagnosis of oral cancer through oral histopathology image analysis. This advancement has significant potential for improving the efficacy of oral cancer treatment strategies.

Deep learning-based auditory attention decoding in listeners with hearing impairment * *This paper’s data analysis is based on a Master’s Thesis work [].

Objective. This study develops a deep learning (DL) method for fast auditory attention decoding (AAD) using electroencephalography (EEG) from listeners with hearing impairment (HI). It addresses three classification tasks: differentiating noise from speech-in-noise, classifying the direction of attended speech (left vs. right) and identifying the activation status of hearing aid noise reduction algorithms (OFF vs. ON). These tasks contribute to our understanding of how hearing technology influences auditory processing in the hearing-impaired population. Approach. Deep convolutional neural network (DCNN) models were designed for each task. Two training strategies were employed to clarify the impact of data splitting on AAD tasks: inter-trial, where the testing set used classification windows from trials that the training set had not seen, and intra-trial, where the testing set used unseen classification windows from trials where other segments were seen during training. The models were evaluated on EEG data from 31 participants with HI, listening to competing talkers amidst background noise. Main results. Using 1 s classification windows, DCNN models achieve accuracy (ACC) of 69.8%, 73.3% and 82.9% and area-under-curve (AUC) of 77.2%, 80.6% and 92.1% for the three tasks respectively on inter-trial strategy. In the intra-trial strategy, they achieved ACC of 87.9%, 80.1% and 97.5%, along with AUC of 94.6%, 89.1%, and 99.8%. Our DCNN models show good performance on short 1 s EEG samples, making them suitable for real-world applications. Conclusion: Our DCNN models successfully addressed three tasks with short 1 s EEG windows from participants with HI, showcasing their potential. While the inter-trial strategy demonstrated promise for assessing AAD, the intra-trial approach yielded inflated results, underscoring the important role of proper data splitting in EEG-based AAD tasks. Significance. Our findings showcase the promising potential of EEG-based tools for assessing auditory attention in clinical contexts and advancing hearing technology, while also promoting further exploration of alternative DL architectures and their potential constraints.

Generation and classification of patch-based land use and land cover dataset in diverse Indian landscapes: a comparative study of machine learning and deep learning models.

In the context of environmental and social applications, the analysis of land use and land cover (LULC) holds immense significance. The growing accessibility of remote sensing (RS) data has led to the development of LULC benchmark datasets, especially pivotal for intricate image classification tasks. This study addresses the scarcity of such benchmark datasets across diverse settings, with a particular focus on the distinctive landscape of India. The study entails the creation of patch-based datasets, consisting of 4000 labelled images spanning four distinct LULC classes derived from Sentinel-2 satellite imagery. For the subsequent classification task, three traditional machine learning (ML) models and three convolutional neural networks (CNNs) were employed. Despite facing several challenges throughout the process of dataset generation and subsequent classification, the CNN models consistently attained an overall accuracy of 90% or more. Notably, one of the ML models stood out with 96% accuracy, surpassing CNNs in this specific context. The study also conducts a comparative analysis of ML models on existing benchmark datasets, revealing higher prediction accuracy when dealing with fewer LULC classes. Thus, the selection of an appropriate model hinges on the given task, available resources, and the necessary trade-offs between performance and efficiency, particularly crucial in resource-constrained settings. The standardized benchmark dataset contributes valuable insights into the relative performance of deep CNN and ML models in LULC classification, providing a comprehensive understanding of their strengths and weaknesses.

The Value of Computer Vision in Identifying the Bone of Dialysis Patients.

In the end stage of kidney disease, abnormal levels of blood calcium, phosphorus, and parathyroid hormone lead to bone metabolism disorders, manifesting as osteoporosis or fibrocystic osteoarthritis. X-ray, CT, and MR are useful for detecting bone lesions in dialysis patients, but currently, computer vision has not yet been used for this purpose. ResNet is a powerful deep CNN model, which has not yet been used to distinguish between the bones of dialysis patients and healthy people. Therefore, this study aimed to investigate the ability of the Resnet50 model to identify the bone of dialysis patients from normal bone. CT images of 200 cases (100 dialysis patients and 100 healthy people aged 31-72 years with male:female ratio of 51:49) were randomly divided into the training and testing groups at the ratio of 8:2. The module of 'torch' was used to train the model of Resnet50 for the current task of image classification. In the test cohort, the accuracy, sensitivity, and specificity with hyper-parameter=0 were 60%, 65%, and 55%, respectively. When the hyper-parameter was 0.6 or 0.7 versus 0, the accuracy was significantly higher (P<0.05). When the hyper-parameter was another number, the accuracy was not significantly different from that with no hyper-parameter (P>0.05). This study has indicated computer vision to be suitable for identifying bone changes caused by dialysis; a hyper-parameter has been found necessary for improving model accuracy. The ResNet50 model with hyper-parameter = 0.7 has exhibited 90% sensitivity in identifying the bone of dialysis patients.

Synthesizing 4D Magnetic Resonance Angiography From 3D Time-of-Flight Using Deep Learning: A Feasibility Study.

Objective and background This study aimed to develop a deep convolutional neural network (DCNN) model capable of generating synthetic 4D magnetic resonance angiography (MRA) from 3D time-of-flight (TOF) images, allowing estimation of temporal changes in arterial flow. TOF MRA provides static information about arterial structures through maximum intensity projection (MIP) processing, but it does not capture the dynamic information of contrast agent circulation, which is lost during MIP processing. Considering the principles of TOF, it is hypothesized that dynamic information about arterial blood flow is latent within TOF signals. Although arterial spin labeling (ASL) can extract dynamic arterial information, ASL MRA has drawbacks, such as longer imaging times and lower spatial resolution than TOF MRA. This study's primary aim is to extend the utility of TOF MRA by training a machine-learning model on paired TOF and ASL data to extract latent dynamic information from TOF signals. Methods A DCNN combining a modified U-Net and a long-short-term memory (LSTM) network was trained on a dataset of 13 subjects (11 menand two women, aged 42-77 years) using paired 3D TOF MRA and 4D ASL MRA images. Subjects had no history of cerebral vessel occlusion or significant stenosis. The dataset was acquired using a 3T MRI system with a 32-channel head coil. Preprocessing involved resampling and intensity normalization of TOF and ASL images, followed by data augmentation and arterial mask generation. The model learned to extract flow information from TOF images and generate 8-phase 4D MRA images. The precision of flow estimation was evaluated using the coefficient of determination (R²) and Bland-Altman analysis. A board-certified neuroradiologist validated the quality of the images and the absence of significant stenosis in the major cerebral arteries. Results The generated 4D MRA images closely resembled the ground-truth ASL MRA data, with R² values of 0.92, 0.85, and 0.84 for the internal carotid artery (ICA), proximal middle cerebral artery (MCA), and distal MCA, respectively. Bland-Altman analysis revealed a systematic error of -0.06, with 95% agreement limits ranging from -0.18 to 0.12. Additionally, the model successfully identified flow abnormalities in a subject with left MCA stenosis, displaying a delayed peak and subsequent flattening distal to the stenosis, indicative of reduced blood flow. Visualization of the predicted arterial flow overlaid on the original TOF MRA images highlighted the spatial progression and dynamics of the flow. Conclusions The DCNN model effectively generated synthetic 4D MRA images from TOF images, demonstrating its potential to estimate temporal changes in arterial flow accurately. This non-invasive technique offers a promising alternative to conventional methods for visualizing and evaluating healthy and pathological flow dynamics. It has significant potential to improve the diagnosis and treatment of cerebrovascular diseases by providing detailed temporal flow information without the need for contrast agents or invasive procedures. The practical implementation of this model could enable the extraction of dynamic cerebral blood flow information from routine brain MRI examinations, contributing to the early diagnosis and management of cerebrovascular disorders.

Early detection of glaucoma: feature visualization with a deep convolutional network

ABSTRACT Computer-aided diagnosis of ocular disorders like glaucoma can be effectively performed with retinal fundus images. Since the advent of machine learning and later deep learning techniques in medical image analysis, research focusing on automated and early detection of glaucoma has gained prominence. In this paper, we show how a deep Convolutional Neural Network model can be used in visualising the features related to Glaucoma that are consistent with (a subset of) those used by ophthalmologists for diagnosis. However, since the doctors do not necessarily depend entirely on these features but back them up with other important measurements such as visual field testing (HVF), any deep neural network such as CNN may not necessarily be expected to yield 100% accuracy. Further, other important parameters such as Intraocular Pressure (IOP), which may be built into a deep learning model, may not necessarily correlate well with the presence of Glaucoma. In this work, a deep CNN model has been trained and tested on a large number of high-quality fundus images, both normal and glaucomatous. We compare the results with transfer learning models such as VGG16, ResNet50, and MobileNetV2. On an average, we obtained an accuracy of 93.75% in identifying glaucoma, focusing only on the features of the fundus image.

Improving the Concrete Crack Detection Process via a Hybrid Visual Transformer Algorithm.

Inspections of concrete bridges across the United States represent a significant commitment of resources, given their biannual mandate for many structures. With a notable number of aging bridges, there is an imperative need to enhance the efficiency of these inspections. This study harnessed the power of computer vision to streamline the inspection process. Our experiment examined the efficacy of a state-of-the-art Visual Transformer (ViT) model combined with distinct image enhancement detector algorithms. We benchmarked against a deep learning Convolutional Neural Network (CNN) model. These models were applied to over 20,000 high-quality images from the Concrete Images for Classification dataset. Traditional crack detection methods often fall short due to their heavy reliance on time and resources. This research pioneers bridge inspection by integrating ViT with diverse image enhancement detectors, significantly improving concrete crack detection accuracy. Notably, a custom-built CNN achieves over 99% accuracy with substantially lower training time than ViT, making it an efficient solution for enhancing safety and resource conservation in infrastructure management. These advancements enhance safety by enabling reliable detection and timely maintenance, but they also align with Industry 4.0 objectives, automating manual inspections, reducing costs, and advancing technological integration in public infrastructure management.

Sign language recognition based on dual-path background erasure convolutional neural network

Sign language is an important way to provide expression information to people with hearing and speaking disabilities. Therefore, sign language recognition has always been a very important research topic. However, many sign language recognition systems currently require complex deep models and rely on expensive sensors, which limits the application scenarios of sign language recognition. To address this issue, based on computer vision, this study proposed a lightweight, dual-path background erasing deep convolutional neural network (DPCNN) model for sign language recognition. The DPCNN consists of two paths. One path is used to learn the overall features, while the other path learns the background features. The background features are gradually subtracted from the overall features to obtain an effective representation of hand features. Then, these features are flatten into a one-dimensional layer, and pass through a fully connected layer with an output unit of 128. Finally, use a fully connected layer with an output unit of 24 as the output layer. Based on the ASL Finger Spelling dataset, the total accuracy and Macro-F1 scores of the proposed method is 99.52% and 0.997, respectively. More importantly, the proposed method can be applied to small terminals, thereby improving the application scenarios of sign language recognition. Through experimental comparison, the dual path background erasure network model proposed in this paper has better generalization ability.

Traffic Flow Optimization at Toll Plaza Using Proactive Deep Learning Strategies

Global urbanization and increasing traffic volume have intensified traffic congestion throughout transportation infrastructure, particularly at toll plazas, highlighting the critical need to implement proactive transportation infrastructure solutions. Traditional toll plaza management approaches, often relying on manual interventions, suffer from inefficiencies that fail to adapt to dynamic traffic flow and are unable to produce preemptive control strategies, resulting in prolonged queues, extended travel times, and adverse environmental effects. This study proposes a proactive traffic control strategy using advanced technologies to combat toll plaza congestion and optimize traffic management. The approach involves deep learning convolutional neural network models (YOLOv7–Deep SORT) for vehicle counting and an extended short-term memory model for short-term arrival rate prediction. When projected arrival rates exceed a threshold, the strategy proactively activates variable speed limits (VSLs) and ramp metering (RM) strategies during peak hours. The novelty of this study lies in its predictive and adaptive capabilities, ensuring efficient traffic flow management. Validated through a case study at Ravi Toll Plaza Lahore using PTV VISSIMv7, the proposed method reduces queue length by 57% and vehicle delays by 47% while cutting fuel consumption and pollutant emissions by 28.4% and 34%, respectively. Additionally, by identifying the limitations of conventional approaches, this study presents a novel framework alongside the proposed strategy to bridge the gap between theory and practice, making it easier for toll plaza operators and transportation authorities to adopt and benefit from advanced traffic management techniques. Ultimately, this study underscores the importance of integrated and proactive traffic control strategies in enhancing traffic management, minimizing congestion, and fostering a more sustainable transportation system.

A Novel Deep Learning Model for Student Performance Prediction Using Engagement Data

Technology-enhanced learning supported by virtual learning environments (VLEs) facilitates tutors and students. VLE platforms contain a wealth of information that can be used to mine insight regarding students’ learning behaviour and relationships between behaviour and academic performance, as well as to model data-driven decision-making. This study introduces a system that we termed ASIST: a novel Attention-aware convolutional Stacked BiLSTM network for student representation learning to predict their performance. ASIST exploits student academic registry, VLE click stream, and midterm continuous assessment information for their behaviour representation learning. ASIST jointly learns the student representation using five behaviour vectors. It processes the four sequential behaviour vectors using a separate stacked bidirectional long short term memory (LSTM) network. A deep convolutional neural network models the diurnal weekly interaction behaviour. It also employs the attention mechanism to assign weight to features based on their importance. Next, five encoded feature vectors are concatenated with the assessment information, and, finally, a softmax layer predicts the high-performer (H), moderate-performer (M), and at-risk (F) categories of students. We evaluate ASIST over three datasets from an Irish university, considering five evaluation metrics. ASIST achieves an area under the curve (AUC) score of 0.86 to 0.90 over the three datasets. It outperforms three baseline deep learning models and four traditional classification models. We also found that the attention mechanism has a slight impact on ASIST’s performance. The ablation analysis reveals that weekly event count has the greatest impact on ASIST, whereas diurnal weekly interaction has the least impact. The early prediction using the first seven weeks of data achieves an AUC of 0.83 up to 0.89 over the three datasets. In yearly analysis, ASIST performs best over the 2018/19 dataset and worst over the 2020/21 dataset.

An Artificial Intelligence Driven Approach for Classification of Ophthalmic Images using Convolutional Neural Network: An Experimental Study.

Early disease detection is emphasized within ophthalmology now more than ever, and as a result, clinicians and innovators turn to deep learning to expedite accurate diagnosis and mitigate treatment delay. Efforts concentrate on the creation of deep learning systems that analyze clinical image data to detect disease-specific features with maximum sensitivity. Moreover, these systems hold promise of early accurate diagnosis and treatment of patients with common progressive diseases. DenseNet, ResNet, and VGG-16 are among a few of the deep learning Convolutional Neural Network (CNN) algorithms that have been introduced and are being investigated for potential application within ophthalmology. In this study, the authors sought to create and evaluate a novel ensembled deep learning CNN model that analyzes a dataset of shuffled retinal color fundus images (RCFIs) from eyes with various ocular disease features (cataract, glaucoma, diabetic retinopathy). Our aim was to determine (1) the relative performance of our finalized model in classifying RCFIs according to disease and (2) the diagnostic potential of the finalized model to serve as a screening test for specific diseases (cataract, glaucoma, diabetic retinopathy) upon presentation of RCFIs with diverse disease manifestations. We found adding convolutional layers to an existing VGG-16 model, which was named as a proposed model in this article that, resulted in significantly increased performance with 98% accuracy (p<0.05), including good diagnostic potential for binary disease detection in cataract, glaucoma, diabetic retinopathy. The proposed model was found to be suitable and accurate for a decision support system in Ophthalmology Clinical Framework.

Explainable DCNN Decision Framework for Breast Lesion Classification from Ultrasound Images Based on Cancer Characteristics.

In recent years, deep convolutional neural networks (DCNNs) have shown promising performance in medical image analysis, including breast lesion classification in 2D ultrasound (US) images. Despite the outstanding performance of DCNN solutions, explaining their decisions remains an open investigation. Yet, the explainability of DCNN models has become essential for healthcare systems to accept and trust the models. This paper presents a novel framework for explaining DCNN classification decisions of lesions in ultrasound images using the saliency maps linking the DCNN decisions to known cancer characteristics in the medical domain. The proposed framework consists of three main phases. First, DCNN models for classification in ultrasound images are built. Next, selected methods for visualization are applied to obtain saliency maps on the input images of the DCNN models. In the final phase, the visualization outputs and domain-known cancer characteristics are mapped. The paper then demonstrates the use of the framework for breast lesion classification from ultrasound images. We first follow the transfer learning approach and build two DCNN models. We then analyze the visualization outputs of the trained DCNN models using the EGrad-CAM and Ablation-CAM methods. We map the DCNN model decisions of benign and malignant lesions through the visualization outputs to the characteristics such as echogenicity, calcification, shape, and margin. A retrospective dataset of 1298 US images collected from different hospitals is used to evaluate the effectiveness of the framework. The test results show that these characteristics contribute differently to the benign and malignant lesions' decisions. Our study provides the foundation for other researchers to explain the DCNN classification decisions of other cancer types.