Methods for object detection, especially those utilizing YOLO (You Only Look Once), are known for their impressive ability to balance precision and speed. However, their potential in detecting brain tumours has received limited attention. RepVGG-GELAN, proposed in this study is an advanced YOLO framework enhanced by RepVGG (Reparameterized Convolutional Neural Network), a specialized convolutional method designed for detection particularly emphasizing brain tumours in medical imaging. With RepVGG architecture the model enhances both speed and accuracy of brain tumour detection. The integration of RepVGG within the YOLO framework seeks to optimize both computational efficiency and detection effectiveness. This research incorporates a Generalized Efficient Layer Aggregation Network (GELAN) architecture based on spatial pyramid pooling enhancing the capabilities of RepVGG. Tests on a brain tumour dataset reveal that RepVGG-GELAN exceeds the performance of the RCS-YOLO (Reparameterized Convolution ShuffleNet YOLO) in terms of both precision and processing speed. Notably, RepVGG-GELAN delivers a 4.91% enhancement in precision and a 2.54% improvement in AP50 (Average Precision at IOU=0.5) over the most recent method with a computational efficiency of 240.7 GFLOPs (Giga Floating Point Operations per Second). The proposed RepVGG-GELAN gives positive results demonstrating itself as a state-of-the-art solution for accurate and efficient detection of brain tumour images. The implementation code is publicly available at https://github.com/ThensiB/RepVGG-GELAN .