Suppr超能文献

用于超分辨率质子磁共振波谱(1H MRS)成像的亚像素移位采集。

Sub-pixel shifted acquisitions for super-resolution proton magnetic resonance spectroscopy (1H MRS) mapping.

作者信息

Abd-Almajeed Adnan, Langevin François

机构信息

AL-Andalus University for Medical Sciences, Bio-Medical Engineering Faculty, Tartous, ALQadmous, Syria.

Université de Technologie de Compiègne (UTC), Division Imagerie Médical, CIMA, BP20.529, 60205 Compiègne cedex, France.

出版信息

Magn Reson Imaging. 2015 May;33(4):448-58. doi: 10.1016/j.mri.2015.01.002. Epub 2015 Jan 16.

Abstract

PURPOSE

(1)H MRS typical studies lead to low-resolution maps (voxels size>1cm(3)). The low-resolution maps may involve the presence of two or more different tissues in a single voxel which gives partial volume effects. Such insufficient spatial resolution remains an obstacle for most routine clinical examinations and restricts the ability in the detection of small lesions.

MATERIALS AND METHODS

The proposed algorithm is based on the combination of three ideas: controlled sub-pixel shift, averaging, and initial pixel size reduction until the spectrum extinction because initial pixel size will determine the final spatial resolution. Frequency range limits are studied to determine the optimal window of spectral peak (WSP). This method is validated in-vitro and in-vivo.

RESULTS

The results in-vitro allowed us to consolidate the theoretical one. Results in-vivo are compared with the clinical protocol to demonstrate the increase in spatial resolution. The proposed method allowed us to reduce the pixel size to1/16 of its initial pixel size in total scan time (TS) of 17min 10s. We demonstrated that the optimal WSP is 80% of its initial value and the interpolation "Bilinear" has the minimum error.

CONCLUSION

A novel approach is established for processing images and reconstructing high-resolution map from low-resolution maps. We have presented a robust algorithm for high spatial resolution in-vivo (1)H MRS, and have achieved a super spatial resolution up to 0.034cm(3) to overcome the major technical challenges.

摘要

目的

(1)氢磁共振波谱(1H MRS)的典型研究得到的是低分辨率图谱(体素大小>1立方厘米)。低分辨率图谱可能在单个体素中包含两种或更多不同组织,从而产生部分容积效应。这种空间分辨率不足仍然是大多数常规临床检查的障碍,并限制了检测小病变的能力。

材料与方法

所提出的算法基于三种理念的结合:可控亚像素移位、平均以及逐步减小初始像素大小直至频谱消失,因为初始像素大小将决定最终的空间分辨率。研究频率范围限制以确定频谱峰值的最佳窗口(WSP)。该方法在体外和体内得到验证。

结果

体外结果使我们能够巩固理论结果。体内结果与临床方案进行比较以证明空间分辨率的提高。所提出的方法使我们能够在17分10秒的总扫描时间内将像素大小减小到其初始像素大小的1/16。我们证明最佳WSP为其初始值的80%,并且“双线性”插值的误差最小。

结论

建立了一种用于处理图像和从低分辨率图谱重建高分辨率图谱的新方法。我们提出了一种用于体内高空间分辨率氢磁共振波谱的稳健算法,并实现了高达0.034立方厘米的超高空间分辨率,以克服主要技术挑战。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验