Gatewood Jason, Peterson Jessica, Aripoli Allison, Huppe Ashley, Winblad Onalisa, Walter Carissa, Young Kate, Li Yanming, Clark Lauren, Deshpande Vibhas, Itriago Pedro, Pinker Katja
Department of Radiology, University of Kansas, Kansas City, KS, USA.
Mercy Hospital, Springfield, MO, USA.
Eur Radiol. 2025 May 7. doi: 10.1007/s00330-025-11624-y.
To determine whether a low b-value (b0) diffusion-weighted imaging (DWI) sequence of a similar acquisition time can fulfill the role of a T2-weighted/fluid-sensitive short tau inversion recovery (STIR) sequence.
In this prospective, single-institutional study, participants underwent 3-T or 1.5-T breast MRI, which included a standard T2-weighted/fluid-sensitive short TI inversion recovery (STIR) sequence and an investigational DWI sequence (comprising readout segmented echo planar imaging (rs-EPI) and simultaneous multi-slice (SMS) imaging). Five radiologists independently scored 121 lesions on imaging by answering three questions assessing lesion conspicuity, lesion signal characteristics, and artifact grading. Statistical analysis included one-sample tests of proportions and Nelson's model kappa.
From 605 assessments in 72 patients (mean age, 48.3 ± 12.5 years [standard deviation]; 72 women), 537 (88.8%, p < 0.001) received assessments that lesion conspicuity on DWI was not inferior to STIR, 586 (96.9%, p < 0.001) received assessments that DWI followed the lesion signal characteristics on STIR, and 595 (98.3%, p < 0.001) received assessments that there were no or mild artifacts on DWI. Inter-reader agreement was substantial for similar lesion conspicuity (89.9%, к = 0.790, 95% CI: 0.76-0.82) and similar lesion signal characteristics (95.7%, к = 0.733, CI 0.69-0.77) on DWI and STIR, and moderate for whether there was mild/no artifact vs. moderate/substantial artifact on DWI (97.2%, к = 0.515, CI: 0.29-0.75).
Low b-value DWI provides similar diagnostic capabilities as T2-weighted/fluid-sensitive STIR, with comparable lesion conspicuity and lesion signal characteristics and without significant artifacts.
Question Can diffusion-weighted imaging (DWI) be incorporated into clinical 1.5-T and 3-T breast MRI protocols without adding acquisition time? Findings The signal characteristics and conspicuity of breast lesions on low b-value (b0) DWI were not inferior compared with that on standard T2-weighted/fluid-sensitive imaging. Clinical relevance The incorporation of low b-value DWI can translate to improved cost-effectiveness and patient experience of breast MRI by reducing false-positives and reducing overall scan time.
确定采集时间相似的低b值（b0）扩散加权成像（DWI）序列是否能起到T2加权/液体敏感短tau反转恢复（STIR）序列的作用。
在这项前瞻性单机构研究中，参与者接受了3-T或1.5-T乳腺MRI检查，其中包括标准的T2加权/液体敏感短TI反转恢复（STIR）序列和一个研究性DWI序列（包括读出分段回波平面成像（rs-EPI）和同时多切片（SMS）成像）。五位放射科医生通过回答三个评估病变清晰度、病变信号特征和伪影分级的问题，对121个病变的影像进行了独立评分。统计分析包括比例的单样本检验和Nelson模型kappa检验。
在72例患者（平均年龄48.3±12.5岁[标准差]；72名女性）中的605次评估中，537次（88.8%，p<0.001）评估认为DWI上的病变清晰度不低于STIR序列，586次（96.9%，p<0.001）评估认为DWI上的病变信号特征与STIR序列上的一致，595次（98.3%，p<0.001）评估认为DWI上无伪影或仅有轻度伪影。对于DWI和STIR上相似的病变清晰度（89.9%，κ=0.790,95%CI：0.76-0.82）和相似的病变信号特征（95.7%，κ=0.733,CI 0.69-0.77）, 读者间一致性较高；对于DWI上是轻度/无伪影还是中度/重度伪影，读者间一致性为中等（97.2%，κ=0.515,CI：0.29-0.75）。
低b值DWI与T2加权/液体敏感STIR具有相似的诊断能力, 病变清晰度和病变信号特征相当，且无明显伪影。
问题：在不增加采集时间的情况下, 扩散加权成像（DWI）能否纳入临床1.5-T和3-T乳腺MRI检查方案？结果：低b值（b0）DWI上乳腺病变的信号特征和清晰度与标准T2加权/液体敏感成像相比并不逊色。临床意义：纳入低b值DWI可通过减少假阳性和缩短整体扫描时间, 提高乳腺MRI的成本效益和患者体验。
