Scholarlink

Early detection of brain tumors plays a significant role in decreasing the rate of mortality due to brain tumors in the patient. The detection of brain tumors has been challenging due to their irregular shapes and diffused boundaries with normal brain tissues. The most popular object-detection algorithm for the lightweight and highly accurate detection of brain tumors has been You Only Look Once (YOLO). This proposed work presents a deep learning-based approach utilizing different MRI brain tumor scans to detect brain tumors from the usual, unaffected area. The proposed model uses YOLOv8 cascaded with the Spatial Attention Module (SAM). The dataset Br35H has been used to validate the proposed model, achieving a 93.6% mean Average Precision (mAP) score for the tumor region. These results demonstrate an improvement compared to existing state-of-the-art (SOTA) techniques.

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

Detection of Brain Tumors Using YOLOv8 Cascaded with Spatial Attention Module

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Savan Kumar,
  • Pranshu Jena,
  • Umesh C. Pati

摘要

Early detection of brain tumors plays a significant role in decreasing the rate of mortality due to brain tumors in the patient. The detection of brain tumors has been challenging due to their irregular shapes and diffused boundaries with normal brain tissues. The most popular object-detection algorithm for the lightweight and highly accurate detection of brain tumors has been You Only Look Once (YOLO). This proposed work presents a deep learning-based approach utilizing different MRI brain tumor scans to detect brain tumors from the usual, unaffected area. The proposed model uses YOLOv8 cascaded with the Spatial Attention Module (SAM). The dataset Br35H has been used to validate the proposed model, achieving a 93.6% mean Average Precision (mAP) score for the tumor region. These results demonstrate an improvement compared to existing state-of-the-art (SOTA) techniques.