English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Ocular Toxoplasmosis (OT), is a common eye infection caused by T. gondii thatcan cause vision problems. Diagnosis is typically done through a clinicalexamination and imaging, but these methods can be complicated and costly,requiring trained personnel. To address this issue, we have created a benchmarkstudy that evaluates the effectiveness of existing pre-trained networks usingtransfer learning techniques to detect OT from fundus images. Furthermore, wehave also analysed the performance of transfer-learning based segmentationnetworks to segment lesions in the images. This research seeks to provide aguide for future researchers looking to utilise DL techniques and develop acheap, automated, easy-to-use, and accurate diagnostic method. We haveperformed in-depth analysis of different feature extraction techniques in orderto find the most optimal one for OT classification and segmentation of lesions.For classification tasks, we have evaluated pre-trained models such as VGG16,MobileNetV2, InceptionV3, ResNet50, and DenseNet121 models. Among them,MobileNetV2 outperformed all other models in terms of Accuracy (Acc), Recall,and F1 Score outperforming the second-best model, InceptionV3 by 0.7% higherAcc. However, DenseNet121 achieved the best result in terms of Precision, whichwas 0.1% higher than MobileNetv2. For the segmentation task, this work hasexploited U-Net architecture. In order to utilize transfer learning the encoderblock of the traditional U-Net was replaced by MobileNetV2, InceptionV3,ResNet34, and VGG16 to evaluate different architectures moreover two differenttwo different loss functions (Dice loss and Jaccard loss) were exploited inorder to find the most optimal one. The MobileNetV2/U-Net outperformed ResNet34by 0.5% and 2.1% in terms of Acc and Dice Score, respectively when Jaccard lossfunction is employed during the training.

Benchmarking Deep Learning Frameworks for Automated Diagnosis of Ocular Toxoplasmosis: A Comprehensive Approach to Classification and Segmentation