Diagnostic and Interventional Radiology, University of Pisa, Via Paradisa, 2 - Cisanello Hospital, 56100, Pisa, Italy.
IMAGO7 Foundation, IRCCS Stella Maris, Pisa, Italy.
Radiol Med. 2022 Sep;127(9):950-959. doi: 10.1007/s11547-022-01533-1. Epub 2022 Aug 19.
To compare the characteristics detected by 7Tesla (7 T) MR and the histological composition of ex-vivo specimens from lesions diagnosed at preoperative CT scan as Pancreatic Ductal Adenocarcinoma (PDAC).
Ten pancreatic specimens were examined. The 7 T imaging protocol included both morphologic and quantitative sequences; the latter was acquired by conventional methods and a novel multiparametric method, the magnetic resonance fingerprinting (MRF) sequence. Two radiologists reviewed the images to: (1) evaluate the quality of the morphological and quantitative sequences by assigning an "image consistency score" on a 4-point scale; (2) identify the lesion, recording its characteristics; (3) perform the quantitative analysis on "target lesion" and "non target tissue". Finally, the specimen was analysed by two pathologists.
Seven out of 10 lesions were PDAC, 2/10 were biliary carcinomas, whereas one lesion was an ampullary adenocarcinoma. The quality of the morphological sequences was judged "excellent". The "image consistency score" for the conventional quantitative sequences and MRF were 2.8 ± 0.42 and 2.9 ± 0.57; the "overall MR examination score" was 3.5 ± 0.53. A statistical correlation was found between the relaxation time values of conventional and MRF T1-weighted sequences (p < 0.0001), as well as between conventional and MRF fat- and water-fraction maps (p < 0.05). The "target lesion" and "non target tissue" relaxation time values were statistically different according to conventional T1-, T2-weighted, and MRF T1-weighted sequences.
Conventional T1-, T2-weighted sequences and MRF derived relaxometries may be useful in differentiating between tumour and non-target pancreatic tissue. Moreover, the MRF sequence can be used to obtain reliable relaxation time data.
比较 7 特斯拉（7T）磁共振（MR）检测到的特征与术前 CT 扫描诊断为胰腺导管腺癌（PDAC）的病变的离体标本的组织学成分。
对 10 个胰腺标本进行了检查。7T 成像方案包括形态学和定量序列；后者通过常规方法和一种新的多参数方法，即磁共振指纹图谱（MRF）序列进行采集。两名放射科医生对图像进行了评估：（1）通过分配 4 分制的“图像一致性评分”来评估形态学和定量序列的质量；（2）识别病变并记录其特征；（3）对“目标病变”和“非目标组织”进行定量分析。最后，两名病理学家对标本进行了分析。
10 个病变中有 7 个为 PDAC，2/10 个为胆管癌，1 个为壶腹腺癌。形态学序列的质量被评为“优秀”。常规定量序列和 MRF 的“图像一致性评分”分别为 2.8±0.42 和 2.9±0.57；“整体 MR 检查评分”为 3.5±0.53。常规 T1 加权序列和 MRF T1 加权序列的弛豫时间值之间存在统计学相关性（p<0.0001），常规和 MRF 脂肪和水分数图之间也存在统计学相关性（p<0.05）。根据常规 T1、T2 加权和 MRF T1 加权序列，“目标病变”和“非目标组织”的弛豫时间值存在统计学差异。
常规 T1、T2 加权序列和 MRF 衍生的弛豫率可用于区分肿瘤和非目标胰腺组织。此外，MRF 序列可用于获得可靠的弛豫时间数据。
