相控阵线圈联合并行采集技术在胃癌扩散加权成像的应用探讨

目的评价相控阵线圈联合并行采集技术（array spatial sensitivity encoding technique,ASSET）消除胃癌磁共振扩散加权成像（diffusion weighted imaging,DWI）伪影及提高图像对比的能力,探讨ASSET技术对表观扩散系数值（apparent diffusion coefficient,ADC）测量的影响。方法12例胃癌患者共计20例次MRI检查入组研究,扩散敏感梯度b=0,1000s/mm^2。根据ASSET的有无设计2组序列,分别为DWIno-ASSET组和DWIASSET组,均采用单次激发EPI序列（SS-EPI）,4次信号平均结合分次屏气技术。评价磁敏感、化学位移、Ghost及ASSET相关伪影在两组序列出现的情况。以肿瘤显示、伪影分级评分、伪影积分、信噪比（signal-to-noiseratio,SNR）及对比噪声比（contrast-to-noise ratio,CNR）为标准,比较2组序列所得图像的质量。比较评价ASSET技术对胃癌癌肿ADC值测量的影响。结果20例次胃癌癌肿在DWIno-ASSET及DWIASSET上均显示为相对高信号。DWIno-ASSET序列15例可见伪影,平均伪影分级评分1.90,平均伪影积分1.10,15%（3/20例）由于明显伪影干扰或图像变形无法进行ADC值测量;DWIASSET序列7例可见伪影,平均伪影分级评分1.35,平均伪影积分0.65,均未干扰ADC值测量。DWIASSET明显消除磁敏感伪影（P〈0.05）,部分消除Ghost伪影（P〉0.05）。两组序列均未见化学位移伪影。7例DWIASSET可见ASSET伪影,均未干扰病变显示及ADC值测量。DWIno-ASSET图像变形较DWIASSET明显,统计学差异有显著性（P〈0.05）。结合ASSET后,图像背景噪声信号强度减低（P〈0.05）,信噪比（SNR）减低（P〈0.05）,对比噪声比（CNR）增高（P〈0.05）。DWIno-ASSET组与DWIASSET组测得癌肿ADC值差异无统计学显著性（P〉0.1）。结论ASSET技术可减轻胃癌DWI成像相关伪影,降低图像噪声信号强度,提高图像对比;结合多信号平均采集提高信噪比,减轻ASSET相关伪影后,可得到较高质量的胃癌扩散加权图像。

Phase array coil combined with parallel imaging in the use of gastric cancer DWI

Objective To investigate the potential of ASSET used to eliminate artifacts and increase contrasts during DWI imaging in patients with gastric cancer (GC), and evaluate the influence of ASSET on the measurements of ADCs. Methods Twenty examinations of 12 gastric cancer patients were enrolled into the study. The b values were 0 and 1000 s/mm2. Two sequences, which were DWI with ASSET (DWIASSET) and without (DWIno-ASSET), used single-shot EPI to generate DWI images. 4NEXs combined with segmented breath-holds were employed. The magnetic susceptibility artifacts, chemical shift artifacts, ghost artifacts, and ASSET-correlating artifacts of 2 sequences were evaluated. The detection rate of GCs, artifacts grade, artifacts score, SNR and CNR were employed as standards to evaluate the quality of DWI images. Evaluate the influence of ASSET on the measurement of ADCs through the images of water phantom and GCs. Results All GCs showed clearly high signals on DWI images in 20 examinations. There were 15 examinations displayed artifacts in DWIno-ASSET, with mean artifacts grade of 1.90 and mean artifacts score of 1.10; 15% (3/20) examinations couldn't measure ADCs because of severe artifacts and/or image distortion. Seven examinations displayed artifacts in DWIASSET with mean artifacts grade of 1.35, and mean artifacts score of 0.65; all examinations could be used to measure ADCs. DWIASSET displayed less magnetic susceptibility artifacts (P<0.05) and ghost artifacts (P>0.05) than DWIno-ASSET, 7 examinations displayed ASSET-correlating artifacts on DWIASSET without interfering lesion demonstration and ADC measurements. Both sequences didn't display chemical shift artifacts. DWIASSET had lower artifact score than DWIno-ASSET with statistical significant (P<0.05). DWIASSET had lower background signal intensity (P<0.05), lower SNR (P<0.05), and higher CNR (P<0.05) than DWIno-ASSET. There was no statistical difference (P>0.1) between DWIno-ASSET and DWIASSET in ADCs. Conclusion ASSET can eliminate artifacts of DWI, decrease back-ground signal intensity, and raise imaging contrast. After combining with multi-NEX to increase the SNR and eliminate ASSET-correlating artifacts, DWIASSET can generate high quality diffusion-weighted images of GCs.