Enhancing Toxoplasmosis Chorioretinitis Detection: A Hybrid ResNet-YOLO Classifier with Fitness Sorted-Shark Smell Optimization

Ophthalmologists often examine fundus pictures to detect toxoplasmosis chorioretinitis, a serious condition of the eye. Preventing blindness in people who may be impacted by this disorder requires prompt and correct diagnosis. In this study, we present a new method for improving Toxoplasmosis chorioretinitis diagnosis accuracy by using a hybrid deep learning model that fuses features from the ResNet and YOLO (You Only Look Once) architectures. The core difference in our method is that we use the YOLO classifier for illness identification instead of the ResNet model's fully connected layer. We use the Fitness Sorted-Shark Smell Optimization (FS-SSO) technique to fine-tune crucial model parameters to get a high recognition rate with this hybrid deep learning model. Using a wide range of performance indicators and accuracy measures, we rigorously assess the effectiveness of the suggested strategy in our experimental investigation. The findings show that our hybrid model outperformed the state-of-the-art methods for detecting Toxoplasmosis chorioretinitis with an accuracy of 99%. Using the FS-SSO approach for parameter optimization, this study demonstrates the promise of combining the strengths of the ResNet and YOLO models for medical picture analysis. Our method's excellent accuracy suggests that it might be used to enhance patient outcomes and minimize the incidence of blindness caused by Toxoplasmosis chorioretinitis by allowing for earlier diagnosis and treatments.