Enhanced Brain Tumor Detection Using Finetuned Transfer Learning Models: Achieving Superior Accuracy with Xception

Diagnosing and identifying brain tumors utilize MRI X-rays and ultrasound along with research-based studies and techniques within medical imaging research. The medical procedure called magnetic resonance imaging (MRI) delivers detailed images as its primary output. Tumor-related problems are an entrance point for numerous machine learning and deep learning approaches that analyze brain MRI data. These methods come with unique disadvantages among them. We applied deep learning transfer learning methods to classify brain images according to this research. The models of Inception and ResNet-50 and VGG-19 and Xception received tuning updates, which brought about substantial performance improvements. The suggested Xception model demonstrated superior performance to other models by reaching 99% accuracy. The proposed model achieved an accuracy of 99% precision, recall, and F-score measure to outperform the sixth-highest model and the state-of-the-art previous work for brain tumor diagnosis. The Xception model shows potential as an appropriate diagnosis instrument for medical use while demonstrating reliable brain tumor prognosis capabilities.