A deep learning approach for detecting Marek’s Disease in poultry combines the YOLOv10 object detection framework with Generative Adversarial Networks (GANs) to enhance the identification of disease specific symptoms, such as leg paralysis, wing drooping, and eye discoloration. YOLOv10’s fast and accurate real-time object detection capabilities allow for the precise localization of these visible markers from images or video footage of poultry. GANs are integrated to generate synthetic images that mimic various stages and presentations of Marek's Disease, addressing the challenges of limited labeled datasets and rare cases. By augmenting the training set with these high-quality synthetic images, the model can generalize better, improving its ability to detect subtle or early stage symptoms of the disease. This method not only increases the model’s robustness in diverse farm environments but also provides an effective tool for early intervention, reducing the spread and economic impact of Marek’s Disease in poultry farming.