RCS-YOLO: A Rapid and Precise Brain Tumor Detection Framework Utilizing the Adam Optimizer

Brain tumors are considered one of the most dangerous health issues, as they can spread quickly and significantly reduce a patient’s chances of survival if not treated promptly. Advanced YOLO frameworks are among the most efficient algorithms for object detection, yet their application in brain tumor detection has been limited. This paper presents a comprehensive strategy for identifying brain tumors using Robo-flow dataset and the existing RCS-YOLO method, enhanced by the RCS-OSA module. We improved accuracy by using the Adam optimizer, incorporating regularization techniques, and increasing the range of 150 to 200 training epochs, it outer-performed the method of Stochastic Gradient Descent. Using Hyperband tuner and the Adam optimizer, we achieved an impressive accuracy of 0.97, along with precision and recall scores of 0.97, indicating that the model generalizes well to unseen data. Our research shows cutting-edge results in the challenge of detecting brain tumors using Robo-flow database.