Scholarlink

Purpose <p>To explore the optimal conditions of neuromelanin imaging for assessing the substantia nigra pars compacta (SNpc) and locus coeruleus (LC) using fast spin echo (FSE) and periodically-rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) sequences, with and without deep learning reconstruction (DLR) technique.</p> Methods <p>Twenty healthy young volunteers underwent FSE and PROPELLER imaging at 3T. Two neuroradiologists scored the signal-to-noise ratio (SNR) of the overall image, as well as contrast and artifacts for each of SNpc and LC, using a 5-point scale. Kappa (κ) coefficients were calculated to evaluate inter-rater agreement. Statistical comparisons of the visual assessment scores among all sequences were performed using Kruskal–Wallis test with post-hoc analysis.</p> Results <p>Inter-rater agreement ranged from fair to perfect. For SNR, both raters’ scores for each sequence with DLR were significantly higher than those without DLR. For SNpc contrast, both raters’ scores for FSE with DLR were significantly higher than those without DLR. For LC contrast and artifact, most scores from both raters for PROPELLER were significantly higher than those for FSE.</p> Conclusion <p>DLR improved image quality and SNpc visualization, and PROPELLER provided superior visualization of the LC in neuromelanin imaging.</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

Visual Assessment of Neuromelanin MR Imaging: A Comparison of PROPELLER and FSE Sequences with and Without Deep Learning Reconstruction

  • 期刊论文

  • 发表日期：

    2025年12月2日

查看全文
  • Azusa Miura,
  • Hiroto Takahashi,
  • Takahiro Nakagawa,
  • Yuki Ito,
  • Arata Otaki,
  • Yumiko Miyauchi,
  • Kana Karasuyama,
  • Chisato Matsuo-Hagiyama,
  • Atsuko Arisawa,
  • Noriyuki Tomiyama

摘要

Purpose

To explore the optimal conditions of neuromelanin imaging for assessing the substantia nigra pars compacta (SNpc) and locus coeruleus (LC) using fast spin echo (FSE) and periodically-rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) sequences, with and without deep learning reconstruction (DLR) technique.

Methods

Twenty healthy young volunteers underwent FSE and PROPELLER imaging at 3T. Two neuroradiologists scored the signal-to-noise ratio (SNR) of the overall image, as well as contrast and artifacts for each of SNpc and LC, using a 5-point scale. Kappa (κ) coefficients were calculated to evaluate inter-rater agreement. Statistical comparisons of the visual assessment scores among all sequences were performed using Kruskal–Wallis test with post-hoc analysis.

Results

Inter-rater agreement ranged from fair to perfect. For SNR, both raters’ scores for each sequence with DLR were significantly higher than those without DLR. For SNpc contrast, both raters’ scores for FSE with DLR were significantly higher than those without DLR. For LC contrast and artifact, most scores from both raters for PROPELLER were significantly higher than those for FSE.

Conclusion

DLR improved image quality and SNpc visualization, and PROPELLER provided superior visualization of the LC in neuromelanin imaging.