Journal: Scientific Reports
Article Title: Optimized deep learning for brain tumor detection: a hybrid approach with attention mechanisms and clinical explainability
doi: 10.1038/s41598-025-04591-3
Figure Lengend Snippet: ( a ) Training and validation loss over 20 epochs. The sharp initial drop followed by convergence in both curves suggests effective learning and minimal overfitting. The model quickly learns to minimize error, reaching stable performance early. ( b ) Training and validation accuracy across epochs. Both curves show a steady rise, with close alignment after epoch 5, indicating strong generalization and consistent performance across unseen MRI data.
Article Snippet: Sandeep Kumar Mathivanan et al. 2024, examined the transfer learning model to detect brain Tumor from the Kaggle Brain MRI data set, and it comprises four transfer learning models ResNet152, VGG19, DenseNet169, MobileNetv3.
Techniques: Biomarker Discovery