<p>PCODs are one of the most prevailing endocrine disorders affecting many women worldwide, and early diagnosis is a critical feature in disease management. In the case of PCOD, ultrasound can present the primary non-invasive platform for visualizing ovarian cysts. Most deep learning-based PCOD automatic detection studies often suffer from generalization of an algorithm in diverse datasets, with low accuracy or poor differentiation capability between cystic and non-cystic structures. In the light of the above issues, we introduce a new deep learning framework called PCODNet-Plus, which is developed to improve the detection and classification of PCOD using ultrasound images. The network is involved in Battle Royale Convolutional Learning Strategy, an optimization strategy that has been boosting the model’s capability for effective feature learning and differentiation, hence enhancing the accuracy of cyst detection as well as improving the overall performance in classification. Extensive experiments on the Kaggle PCOD ultrasound dataset have documented the efficacy of PCODNet-Plus over traditional machine learning models in terms of higher accuracy, precision, recall, and F1-score. The findings indicate that PCODNet-Plus is an effective and reliable framework for the accurate diagnosis of PCOD, promising a robust solution for clinicians and medical experts to automate the detection of PCOD for better patient outcomes.</p>