Scholarlink

In the healthcare field, Magnetic Resonance Imaging (MRI) is widely used in clinics to observe and assess brain tumors, aiding diagnosis and treatment. However, classifying multi-class brain tumors from MRI data is challenging due to the varying size and shape of tumors. To address this, a modified Deep Learning (DL)-based UNet-NASNetLarge model with Random Forest (RF) is proposed for predicting and classifying tumor locations in the brain. The UNet-encoder extracts features from MRI scans, while the NASNet decoder selects feature maps for faster training. Min-max normalization enhances color strength for specific feature boundaries, and the RF technique ensures accurate tumor classification. This system has been validated on a brain tumor dataset, showing superior performance compared to existing state-of-the-art (SOTA) classification methods.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Automated Multi-class Brain Tumor Classification Using Deep Learning-Based Network for MRI Dataset

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Anita Murmu,
  • Sangeeta Kumari,
  • Rajiv Murmu

摘要

In the healthcare field, Magnetic Resonance Imaging (MRI) is widely used in clinics to observe and assess brain tumors, aiding diagnosis and treatment. However, classifying multi-class brain tumors from MRI data is challenging due to the varying size and shape of tumors. To address this, a modified Deep Learning (DL)-based UNet-NASNetLarge model with Random Forest (RF) is proposed for predicting and classifying tumor locations in the brain. The UNet-encoder extracts features from MRI scans, while the NASNet decoder selects feature maps for faster training. Min-max normalization enhances color strength for specific feature boundaries, and the RF technique ensures accurate tumor classification. This system has been validated on a brain tumor dataset, showing superior performance compared to existing state-of-the-art (SOTA) classification methods.