Scholarlink

<p>Brain tumor detection and classification is a critical task in medical imaging aimed at improving diagnostic accuracy and treatment outcomes. This study presents an advanced methodology for brain tumor detection and classification using MRI and X-ray images, leveraging the ADKF-SSECT-SAGAN-NGOA model. The proposed method begins with preprocessing with anisotropic diffusion and Kuwahara filtering (ADKF), which is a hybrid filtering technique that combines anisotropic diffusion and Kuwahara filtering to reduce noise while maintaining edges and improving the local contrast. The proposed algorithm performs feature extraction using the aynchro squeezing extract chirplet transform (SSECT), followed by segmentation with SegNet. The proposed model integrates self-attention mechanisms within generative adversarial networks (GANs) to enhance image resolution and classification accuracy. Additionally, the Northern Goshawk optimization algorithm (NGOA) was employed to optimize the model parameters effectively. The proposed algorithm demonstrates the model's superior performance, achieving an average accuracy of 99.45% for the magnetic resonance imaging (MRI) dataset (a), 99.71% for the MRI dataset (b), and 99.1% for X-ray images. The model also shows remarkable efficiency, with significantly reduced computation times compared to existing methods. This combination of high accuracy and efficiency underscores the potential of the proposed model as a reliable tool for medical image analysis and clinical diagnostics.</p>

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

Efficient and accurate brain tumor detection and classification using advanced hybrid filtering and self-attention generative adversarial networks

  • 期刊论文

  • 发表日期：

    2025年5月7日

查看全文
  • R. Rajakumari,
  • A. Selvapandian

摘要

Brain tumor detection and classification is a critical task in medical imaging aimed at improving diagnostic accuracy and treatment outcomes. This study presents an advanced methodology for brain tumor detection and classification using MRI and X-ray images, leveraging the ADKF-SSECT-SAGAN-NGOA model. The proposed method begins with preprocessing with anisotropic diffusion and Kuwahara filtering (ADKF), which is a hybrid filtering technique that combines anisotropic diffusion and Kuwahara filtering to reduce noise while maintaining edges and improving the local contrast. The proposed algorithm performs feature extraction using the aynchro squeezing extract chirplet transform (SSECT), followed by segmentation with SegNet. The proposed model integrates self-attention mechanisms within generative adversarial networks (GANs) to enhance image resolution and classification accuracy. Additionally, the Northern Goshawk optimization algorithm (NGOA) was employed to optimize the model parameters effectively. The proposed algorithm demonstrates the model's superior performance, achieving an average accuracy of 99.45% for the magnetic resonance imaging (MRI) dataset (a), 99.71% for the MRI dataset (b), and 99.1% for X-ray images. The model also shows remarkable efficiency, with significantly reduced computation times compared to existing methods. This combination of high accuracy and efficiency underscores the potential of the proposed model as a reliable tool for medical image analysis and clinical diagnostics.