Anti-motion Ultrafast T 2 Mapping Technique for Quantitative Detection of the Normal-Appearing Corticospinal Tract Changes in Subacute-Chronic Stroke Patients with Distal Lesions
Yue Zhang,Xiao Wang,Ming Ye,Zongye Li,Yuchuan Zhuang,Qinqin Yang,Qichang Fu,Rui Chen,Eryuan Gao,Yanan Ren,Yong Zhang,Shuhui Cai,Zhong Chen,Congbo Cai,Yanbo Dong,Jianfeng Bao,Jingliang Cheng
DOI: https://doi.org/10.1016/j.acra.2023.11.036
IF: 5.482
2023-12-24
Academic Radiology
Abstract:Rationale and Objectives Stroke patients commonly face challenges during magnetic resonance imaging (MRI) examinations due to involuntary movements. This study aims to overcome these challenges by utilizing multiple overlapping-echo detachment (MOLED) quantitative technology. Through this technology, we also seek to detect microstructural changes of the normal-appearing corticospinal tract (NA-CST) in subacute-chronic stroke patients. Materials and Methods 79 patients underwent 3.0 T MRI scans, including routine scans and MOLED technique. A deep learning network was utilized for image reconstruction, and the accuracy, reliability, and resistance to motion of the MOLED technique were validated on phantoms and volunteers. Subsequently, we assessed motor dysfunction severity, ischemic lesion volume, T 2 values of the bilateral NA-CST, and the T 2 ratio (rT 2 ) between the ipsilesional and contralesional NA-CST in patients. Results The MOLED technique showed high accuracy ( P < 0.001) and excellent repeatability, with a mean coefficient of variation (CoV) of 1.11%. It provided reliable quantitative results even under head movement, with a mean difference (Mean diff ) = 0.28% and a standard deviation difference (SD diff ) = 1.34%. Additionally, the T 2 value of the ipsilesional NA-CST was significantly higher than contralesional side ( P < 0.001), and a positive correlation was observed between rT 2 and the severity of motor dysfunction (r s = 0.575, P < 0.001). Furthermore, rT 2 successfully predicted post-stroke motor impairment, with an area under the curve (AUC) was 0.883. Conclusion The MOLED technique offers significant advantages for quantitatively imaging stroke patients with involuntary movements. Additionally, T 2 mapping from MOLED can detect microstructural changes in the NA-CST, potentially aiding in monitoring stroke-induced motor impairment.
radiology, nuclear medicine & medical imaging