Scholarlink

The Swoop system of Hyperfine Inc. is an affordable, ultra-low-field MRI developed for use in a clinical setting. However, despite its advantages, the relatively low resolution of 64mT MRI data poses additional challenges, especially in examining small structures such as the hippocampus or vessels. As a part of our attempt at the Low field pediatric brain magnetic resonance Image Segmentation and Quality Assurance (LISA) Challenge 2024, we developed two deep learning-based models. First, to evaluate the image quality of 64mT T2 brain MRI data, we implemented an axis classifier module to improve the model performance. Second, for segmentation of the hippocampus in the MRI, a multi-label learning method was used for more accurate segmentation. With these models, we expect to alleviate the accessibility barrier to brain MRI.

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

Axis-Guided Quality Assessment and Multi-label Hippocampal and Ventricular Segmentation in Low-Resolution Pediatric Brain MRI

  • 会议论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年

查看全文
  • Hyunwook Kim,
  • Jinew Seo,
  • Seiyoung Ryu,
  • Joon hyung Park,
  • Sungchul On,
  • Jinwha Choi

摘要

The Swoop system of Hyperfine Inc. is an affordable, ultra-low-field MRI developed for use in a clinical setting. However, despite its advantages, the relatively low resolution of 64mT MRI data poses additional challenges, especially in examining small structures such as the hippocampus or vessels. As a part of our attempt at the Low field pediatric brain magnetic resonance Image Segmentation and Quality Assurance (LISA) Challenge 2024, we developed two deep learning-based models. First, to evaluate the image quality of 64mT T2 brain MRI data, we implemented an axis classifier module to improve the model performance. Second, for segmentation of the hippocampus in the MRI, a multi-label learning method was used for more accurate segmentation. With these models, we expect to alleviate the accessibility barrier to brain MRI.