Deep Ensemble Learning for Retinal Image Classification

Ho, Edward; Wang, Edward; Youn, Saerom; Sivajohan, Asaanth; Lane, Kevin; Jin, Chun; Hutnik, Cindy

doi:10.1167/tvst.11.10.39

Cited by 17 publications

(7 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fundus images are used in the automated diagnosis of multiple ophthalmic diseases based on the SE-ResNeXT model. Ho et al evaluated the model in the Retinal Fundus Multi-Disease Image Dataset (RFMiD) and achieved an ROC area of 0.9586 [ 37 ]. Based on an SE-ResNeXT network, Li et al classified 12 different retinal pathologies based on color fundus images and obtained an AUROC of 0.95 [ 38 ].…”

Section: Resultsmentioning

confidence: 99%

Automated Computer-Aided Detection and Classification of Intracranial Hemorrhage Using Ensemble Deep Learning Techniques

Umapathy,

Murugappan,

Bharathi

et al. 2023

Diagnostics

View full text Add to dashboard Cite

Diagnosing Intracranial Hemorrhage (ICH) at an early stage is difficult since it affects the blood vessels in the brain, often resulting in death. We propose an ensemble of Convolutional Neural Networks (CNNs) combining Squeeze and Excitation–based Residual Networks with the next dimension (SE-ResNeXT) and Long Short-Term Memory (LSTM) Networks in order to address this issue. This research work primarily used data from the Radiological Society of North America (RSNA) brain CT hemorrhage challenge dataset and the CQ500 dataset. Preprocessing and data augmentation are performed using the windowing technique in the proposed work. The ICH is then classified using ensembled CNN techniques after being preprocessed, followed by feature extraction in an automatic manner. ICH is classified into the following five types: epidural, intraventricular, subarachnoid, intra-parenchymal, and subdural. A gradient-weighted Class Activation Mapping method (Grad-CAM) is used for identifying the region of interest in an ICH image. A number of performance measures are used to compare the experimental results with various state-of-the-art algorithms. By achieving 99.79% accuracy with an F-score of 0.97, the proposed model proved its efficacy in detecting ICH compared to other deep learning models. The proposed ensembled model can classify epidural, intraventricular, subarachnoid, intra-parenchymal, and subdural hemorrhages with an accuracy of 99.89%, 99.65%, 98%, 99.75%, and 99.88%. Simulation results indicate that the suggested approach can categorize a variety of intracranial bleeding types. By implementing the ensemble deep learning technique using the SE-ResNeXT and LSTM models, we achieved significant classification accuracy and AUC scores.

show abstract

Section: Resultsmentioning

confidence: 99%

Automated Computer-Aided Detection and Classification of Intracranial Hemorrhage Using Ensemble Deep Learning Techniques

Umapathy,

Murugappan,

Bharathi

et al. 2023

Diagnostics

View full text Add to dashboard Cite

show abstract

“…Convolutional Neural Networks (CNNs), a subset of DL, are revolutionising image recognition capabilities, including in retinal pathology diagnosis [119]. Recent advancements also highlight the application of AI in analysing OCT images, offering the potential for precise disease stage identification and treatment customisation [120].…”

Section: Big Data Internet Of Things (Iot) and Aimentioning

confidence: 99%

Teleophthalmology in Retinal Diseases

El-Badawi,

Goodchild,

Drukarch

et al. 2024

Biomedical Engineering

View full text Add to dashboard Cite

Recent advancements in teleophthalmology have transformed retinal disease management, benefiting healthcare providers and patients. By enabling remote monitoring, teleophthalmology significantly reduces the need for in-person consultations. Easy-to-use devices like at-home Optical Coherence Tomography (OCT) systems empower patients to generate high-quality images crucial for tailored treatment. Artificial intelligence (AI) aids in quick, affordable screenings by analysing fundus photographs and OCT images. These innovations underscore teleophthalmology’s pivotal role in streamlining patient care and optimising healthcare resources. Legal, ethical, and logistical considerations surrounding teleophthalmology, drawn from literature and experiences at Bristol Eye Hospital, are also discussed.

show abstract

“…Ensemble learning includes various learning models to achieve better predictive performance than a single model. Ensemble methodologies are broadly classified as Homogeneous Ensemble approaches, involving Bagging and Boosting, Heterogenous Ensemble approaches involving Stacking, and Majority voting algorithms [52]- [55].…”

Section: Ensemble Learningmentioning

confidence: 99%

Unveiling visionary frontiers: a survey of cutting-edge techniques in deep learning for retinal disease diagnosis

Rajatha,

Davanageri Virupakshappa

2024

IJEECS

View full text Add to dashboard Cite

Retinal disorders impact millions of people globally. These disorders can be detected and diagnosed early enough to not only cure but also avoid permanent blindness. Manual identification of these diseases has always been tedious, time-consuming, and inconsistent. For ophthalmologists, retinal fundus images are a valuable source of information in diagnosing retinal diseases. Automatic identification of eye disorders using artificial intelligence (AI) based learning models has seen substantial development in the computer vision sector recently. Various models, particularly deep learning (DL) models are incredible in identifying and classifying diseases. In the presented review, we have performed an in-depth analysis of various existing DL models, involving preprocessing, classification, segmentation, and techniques to deal with data imbalance. We have also endeavored to gauge the effectiveness of these models by evaluating their performance using the metrics employed in their assessment. In addition, we explored various challenges along with the potential future scope in this domain.<p> </p>

show abstract

Deep Ensemble Learning for Retinal Image Classification

Cited by 17 publications

References 21 publications

Automated Computer-Aided Detection and Classification of Intracranial Hemorrhage Using Ensemble Deep Learning Techniques

Automated Computer-Aided Detection and Classification of Intracranial Hemorrhage Using Ensemble Deep Learning Techniques

Teleophthalmology in Retinal Diseases

Unveiling visionary frontiers: a survey of cutting-edge techniques in deep learning for retinal disease diagnosis

Contact Info

Product

Resources

About