An efficient deep learning model to categorize brain tumor using reconstruction and fine-tuning

• DOI: 10.1016/j.eswa.2023.120534
• Publication Date: 2023-05-27T01:06:53Z
• Authors: Md. Alamin Talukder; Md. Manowarul Islam; Md. Ashraf Uddin; Arnisha Akhter; Md. Alamgir Jalil...; Sunil Aryal; Muhammad Ali Abdu...; Khondokar Fida Hasan; Mohammad Ali Moni
• Keywords: Fine-tuning; Brain tumor

Brain tumors are among the most fatal and devastating diseases, often resulting in significantly reduced life expectancy. An accurate diagnosis of brain is crucial to devise treatment plans that can extend lives affected individuals. Manually identifying analyzing large volumes MRI data both challenging time-consuming. Consequently, there a pressing need for reliable deep learning (DL) model accurately diagnose tumors. In this study, we propose novel DL approach based on transfer effectively classify Our method incorporates extensive pre-processing, architecture reconstruction, fine-tuning. We employ several algorithms, including Xception, ResNet50V2, InceptionResNetV2, DenseNet201. experiments used Figshare tumor dataset, comprising 3,064 images, achieved accuracy scores 99.40%, 99.68%, 99.36%, 98.72% DenseNet201, respectively. findings reveal ResNet50V2 achieves highest rate 99.68% outperforming existing models. Therefore, our proposed model's ability short timeframe aid neurologists clinicians making prompt precise diagnostic decisions patients.