Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada.
Schulich Heart Research Program and Physical Sciences Platform, Sunnybrook Research Institute, Toronto, Ontario, Canada.
Magn Reson Med. 2018 Jun;79(6):3018-3031. doi: 10.1002/mrm.26970. Epub 2017 Oct 13.
To achieve consistent effectiveness in reconstruction of fine image features for cases of varying contrast-to-noise ratio (CNR) to facilitate translating accelerated multicontrast volumetric imaging with isotropic resolution toward clinical utility in peri-infarct characterization.
A low-rank and spatially varying edge-preserving constrained compressed sensing parallel imaging reconstruction method (CP-LASER) is developed to effectively preserve contrast of small-scale structures for highly accelerated multicontrast volumetric imaging in CNR-limited scenarios. CP-LASER synergistically integrates parallel imaging, low-rank and spatially varying edge-preserving sparse modeling to achieve high signal-to-noise-ratio efficiency by leveraging prior knowledge about signal properties including coil sensitivity weighting, spatiotemporally correlated signal relaxation, and spatially varying sparsity.
In the preclinical study using highly accelerated multicontrast volumetric imaging with an isotropic 1.5-mm resolution, CP-LASER demonstrated robust multicontrast reconstruction of peri-infarct characteristics with excellent correspondence with histopathology. CP-LASER provides better delineation of the peri-infarct border zone with improved sharpness than alternative methods in a clinical demonstration on 1.5T with an isotropic 2.2-mm resolution achieved in a single breath-hold.
Accelerated multicontrast volumetric imaging with isotropic resolution using CP-LASER has demonstrated the potential to improve peri-infarct characterization in a clinical setting. Magn Reson Med 79:3018-3031, 2018. © 2017 International Society for Magnetic Resonance in Medicine.
实现针对不同对比噪声比（CNR）情况的精细图像特征重建效果的一致性，以促进各向同性分辨率的加速多对比度容积成像向在梗死周边区特征描述中的临床应用转化。
开发了一种低秩和空间变化边缘保持约束压缩感知并行成像重建方法（CP-LASER），以在 CNR 受限情况下有效地保持小尺度结构的对比度，实现高加速多对比度容积成像。CP-LASER 协同集成并行成像、低秩和空间变化边缘保持稀疏建模，通过利用关于信号特性的先验知识，包括线圈灵敏度加权、时空相关信号弛豫和空间变化稀疏性，实现高信噪比效率。
在使用各向同性 1.5mm 分辨率的高度加速多对比度容积成像的临床前研究中，CP-LASER 对梗死周边区特征进行了稳健的多对比度重建，与组织病理学具有极好的对应性。CP-LASER 提供了比替代方法更好的梗死周边边界带描绘，在 1.5T 上实现了单次屏气 2.2mm 各向同性分辨率的临床演示中，其锐度得到了改善。
使用 CP-LASER 的各向同性分辨率加速多对比度容积成像已显示出在临床环境中改善梗死周边区特征描述的潜力。磁共振医学 79：3018-3031，2018。©2017 国际磁共振学会。
