分享:
分享到微信朋友圈
X
技术研究
乳腺1H磁共振波谱

张卫军,付彩霞.乳腺1H磁共振波谱.磁共振成像, 2011, 2(4): 296-299. DOI:10.3969/j.issn.1674-8034.2011.04.011.


[摘要] 对比增强MRI已经成为重要的乳腺成像手段,但是该手段并不总能提供确切的病理。在动态增强MRI基础上增加活体MRS的初步研究显示的结果很具前景,越来越多的研究小组将MRS加入乳腺MR检查序列。本文阐明了进行MRS检查的预期检查结果,并列举了其中的一些缺陷。
[Abstract] Contrast-enhanced MRI has gained acceptance as an important breast imaging modality. However it does not always provide a definitive pathology. Initial studies where in vivo proton MR spectroscopy has been added as an adjunct to dynamic contrast-enhanced MR imaging of the breast have shown promising results and a growing number of research groups are incorporating the technique into their breast MR protocols. The aim of this article is to illustrate the expected examination results and outline some of the pitfalls associated with undertaking a breast MRS examination.
[关键词] 磁共振成像;磁共振波谱学;乳腺
[Keywords] Magnetic resonance imaging;Magnetic resonance spectroscopy;Breast

通讯作者:张卫军,E-mail: weijun.zhang@siemens.com


第一作者简介:
        张卫军(1974-),女,医学博士,西门子(深圳)磁共振有限公司MR应用专家。

收稿日期:2011-06-05
接受日期:2011-07-08
中图分类号:R445.2; R737.9 
文献标识码:A
DOI: 10.3969/j.issn.1674-8034.2011.04.011
张卫军,付彩霞.乳腺1H磁共振波谱.磁共振成像, 2011, 2(4): 296-299. DOI:10.3969/j.issn.1674-8034.2011.04.011.

0 引言

       本文原英文版刊登于西门子MAGNETOM Flash杂志2010年第一期的ISMRM特刊第32~42页,题目为:Proton Magnetic Resonance Spectroscopy of the Breast,由哈佛医学院布莱汉姆妇女医院的Peter Stanwell博士(现工作于澳大利亚纽卡斯尔大学)等6位专家经过长期对乳腺磁共振波谱(magnetic resonance spectroscopy, MRS)的临床研究撰写而成。译者认为,该文通过列举大量的临床数据,对乳腺MRS的优势及存在的问题进行了深入探讨,具有很高的学术价值,故将其翻译成中文,与国内的MRI临床专家们分享。

       对比增强MRI已经成为重要的乳腺成像手段,但是该手段并不总能提供确切的病理[1]。2007年Lehman[2]报告MRI在乳腺检查以及乳腺X线摄影等初步检查诊断单侧乳腺癌后,可以快速探测对侧乳腺癌,以提高诊断水准。该研究发现在该组病人中,MRI的特异性是88%。随后的美国癌症协会关于乳腺筛查指南将MRI作为乳腺X线摄影的辅助手段,称:"MRI扫描较乳腺摄片更敏感,但癌症病灶或良性病灶都更易显示。如果接下来不进行活检或其他有创干预,通常没有办法判断这些病灶是否为恶性"[1]。现在的问题是:"增加1H MRS检查,是否可以增加MRI的精确性?"

       活体MRS可以与常规MRI检查一块进行,得到乳腺或者乳腺病灶化学成分的信息。在动态增强MRI基础上增加活体MRS的初步研究显示的结果很具前景,越来越多的研究小组将MRS加入乳腺MR扫描协议。该论文就是要阐明进行MRS检查的预期检查结果,并列举出其中的一些缺陷。初期,在1.5 T上,用MRS诊断乳腺癌是基于在3.2 ppm处观察含胆碱的代谢物(total Choline, tCho)。观察到的tCho信号的共振频率用来鉴别正常的腺体组织和恶性组织[3]。同样地,观察到的tCho信号的定量用来作为鉴别良性和恶性病灶的手段[4],还可用来监测化疗反应[5]

1 乳腺MRS的应用进展

1.1 tCho与位于3.23 ppm和3.28 ppm分辨清晰的共振峰

       以前测量tCho,是因为不能达到区分共振频率所必需的波谱分辨率。然而,识别tCho信号确切的共振频率可能对更精确地区分良恶性更有帮助。而且,对于监测化疗也更有帮助,因为化疗后有些成分没有发生改变,而另一些成分的共振峰可能发生了改变[6]。目前的硬件在处理数据后可以探测并且比较tCho及各胆碱共振频率(图1[3,7]。为了进行比较,需要在实验方面进行许多考虑。

图1  乳腺活体单体素波谱(3 T, PRESS, TR=2000 ms, TE=135 ms,平均192次)。波谱配准参考1.33 ppm处的脂肪的亚甲基峰和4.74 ppm处的水峰。上面的波谱来自于乳腺癌患者,在3.23 ppm处可见共振峰,符合磷酸胆碱。下面为纤维腺瘤波谱作为对照。tChol区为总胆碱
Fig 1  In vivo breast single voxel spectra (3 T, PRESS, TE=135 ms / TR=2000 ms, 192 signal averages). Spectra are processed as described were referenced to the methylene resonance of lipid at 1.33 ppm and water at 4.74 ppm. In the spectrum from the cancer bearing patient (top) the resonance is at 3.23 ppm and consistent with phosphocholine. For comparison, a spectrum derived from a fibroadenoma (bottom) is shown. The total choline "tChol" region is shown in blue.

1.2 无微浸润的乳管原位癌(DCIS)

       我们的经验中,一小部分病理是乳腺导管原位癌(ductal carcimona in situ, DCIS)的无微浸润的病例,共振峰显示在3.28 ppm。见图2

图2  乳腺活体单体素波谱(1.5 T, TR=2000 ms, TE=135 ms, 256次平均),病灶术中证实为无微浸润的原位癌(DCIS)。本图数据处理来自图1
Fig 2  In vivo breast single voxel spectrum (1.5 T; TR=135 ms; TE=2000 ms; 256 averages) of a lesion identified intraoperatively as DCIS with no microinvasion. Data was processed and referenced as shown in figure 1.

1.3 可探测病灶大小

       最新的MR技术和线圈使我们可以对大于等于8 mm的需要进行病理判断的病灶进行MRS分析。对活体探测到的胆碱峰的意义有更深的理解,特别是在这个波谱的区域探测到波峰,有可能为"正常"或者浸润前的乳腺组织,有可能探测到非恶性的其他病变,如激素性的。我们和其他一些小组已经证实1H MRS能够成功地在术前,无创活体应用于乳腺,在1 cm或者更大的肿瘤诊断出乳腺病理为恶性、良性或正常组织,精确性较高。

2 波谱采集的影响因素

2.1 实验注意事项

       该技术在作者的医院常规应用,操作者均受到严格训练。由于可能遇到许多技术问题,要求有对扫描仪的专业知识。下面我们总结有可能碰到到一些问题,以帮助大家获得高质量的波谱数据。

       2.1.1 用MRI扫描设备上的标准软件可能得到什么样的结果?见图3图4

       2.1.2 匀场(见图5

       2.1.3 参考(见图6

       2.1.4 定量(见图7

       2.1.5 脂肪

       乳腺活体波谱常受脂肪峰的影响。当脂肪组织在体素内但不构成病变的成分时可能导致若干问题[7,8]。可以通过延长TE[9]或者采用平均TE采集来降低脂肪的影响[8]。见图8图9

       (未完待续)

图3  正常志愿者的乳腺组织,1H单体素波谱,PRESS序列采集,TR:2000ms,TE:135ms。图3A.未压水波谱示很强的水、脂信号,胆碱峰几乎难以观察。图3B.纵轴放大后,显示胆碱峰信号
Fig 3  Proton single-voxel spectrum acquired with the PRESS sequence (repetition time (TR) 2000 ms, echo time (TE) 135 ms) from glandular tissue in a breast volunteer. Fig 3A. Non-water suppressed spectrum shows strong water and lipid signals and a barely visible choline resonance (arrow). Fig 3B. The vertical display is increased with the choline signal becoming more apparent.
图4  活检证实为乳腺癌,1H单体素波谱,PRESS序列采集,TR:2000 ms,TE:135 ms。图4A.未压水波谱示很强的水、脂信号,胆碱峰几乎难以观察。图4B.纵轴放大后,显示胆碱峰信号
Fig 4  Proton single-voxel spectrum acquired with the PRESS sequence (repetition time (TR) 2000 ms, echo time (TE) 135 ms) from a biopsy-proven breast cancer. Fig 4A. Non-water suppressed spectrum shows strong water and lipid signals and a barely visible choline resonance (arrow). Fig 4B. The vertical display is increased with the choline signal becoming more apparent.
图5  体素匀场;正常志愿者的乳腺组织,1H单体素波谱,PRESS序列采集,TR: 2000 ms,TE: 135 ms 5A:运行自动匀场程序得到的波谱。在0.9和1.3 ppm处可见明显的脂峰;5B:手动匀场采集的波谱,仍旧可以在0.9和1.3 ppm处看见明显的脂峰,但是谱的分辨率增加,在基线上可以分辨两个波峰
Fig 5  Voxel shimming. Proton single-voxel spectrum acquired with the PRESS sequence (repetition time (TR) 2000 ms, echo time (TE) 135 ms) in a breast volunteer. 5A: Spectrum was acquired using the automatic shimming program only. Prominent signals are visible from lipids at 0.9 and 1.3 ppm. 5B: Spectrum was acquired with the addition of manual shimming. Prominent signals from lipids are again noted (0.9 & 1.3 ppm) but with enhanced spectral resolution with these two peaks now resolved to their baselines.
图6  正确的参考在未压水的波谱上进行,可以提供区分3.23 ppm和3.28 ppm峰的必要的精确性
Fig 6  Correct referencing is undertaken on the non suppressed water spectrum and provides necessary accuracy which is required to distinguish the 3.23 ppm from 3.28 ppm resonances seen in figures 1 and 2.
图7  t-含胆碱化合物的定量。体素的定位显示在左侧;压水波谱显示在右侧,拟和的tCho峰显示在上方,实际峰和拟和峰的差显示在下方。第一排:正常腺体,体积13ml,[tCho]=0.66±0.06 mmol/kg,脂肪含量=3%;第二排:侵袭性乳管癌,体积6.8 ml, [tCho] = 6.1±0.08 mmol/kg,脂肪含量= 8%;第三排:不典型增生,体积1.1 ml, [tCho]=1.4±0.7 mmol/kg,脂肪含量= 14% (引自Bolan等)[4]
Fig 7  Quantification of t-choline-containing compounds. For each row the voxel location is shown on a contrast-enhanced localizer image on the left, while a water-suppressed spectrum is shown on the right, with the tCho fit shown above and the residual underneath. Top row, normal gland, volume = 13.0 ml, [tCho] = 0.66±0.06 mmol/kg, lipid fraction = 3%. Middle row, malignant tumor of invasive ductal carcinoma, volume=6.8 ml, [tCho]=6.1±0.08 mmol/kg, lipid fraction= 8%. Bottom row, atypical hyperplasia volume = 1.1 ml, [tCho] =1.4 ±0.7 mmol/kg, lipid fraction = 14%. Adapted from Bolan, et al.[4]
图8  TE对乳腺波谱的影响。乳腺癌经活检证实,采用PRESS序列获得1H单体素波谱。8A:显示体素在乳腺癌的位置;8B:单体素波谱(TR: 2000 ms, TE:135 ms);8C:单体素波谱(TR: 2000 ms,TE: 270 ms)。将TE从135升至270 ms,由于T2衰减增加,脂肪的幅度下降(位于0.90和1.33 ppm)[7]
Fig 8  Effect of echo time (TE) in breast spectroscopy. Proton single-voxel spectrum acquired with the PRESS sequence from a biopsy-proven breast cancer. 8A: Indicates the voxel placement within the breast cancer. 8B: Single-voxel spectrum acquired with TR of 2000 ms, TE of 135 ms. 8C: Single-voxel spectrum acquired with TR of 2000 ms, TE of 270 ms. Increasing the TE from 135 to 270 ms reduces the amplitude of lipid (at 0.90 and 1.33 ppm) due to increased T2 relaxation[7].
图9  脂肪导致的边带。乳腺癌患者采用1H单体素波谱。9A:对比增强图像,15 mm×16 mm×15 mm大小体素位于一活检证实的炎性乳腺癌,包含部分脂肪组织。9B:压水波谱。上面的波谱所用的TE为45 ms,下面的波谱为平均TE (TE 45~196 ms)采集。胆碱信号在平均TE法更加明显,由于梯度导致的伪影较单TE采集减少(引自Bolan等[8]
Fig 9  Lipid-induced sidebands. Proton single-voxel spectrum collected in a breast cancer patient. 9A: Contrast-enhanced image demonstrates a 15 mm×16 mm×15 mm voxel surrounding a biopsy-proven case of inflammatory breast cancer with included surrounding adipose tissue. 9B: Water-suppressed spectra, top spectrum collected with a single echo time (45 ms), bottom spectrum collected with TE-averaged acquisition (TE 45-196 ms). The Cho signal (vertical arrows) is more distinct in the TE-averaged acquisition due to reduction in gradient-induced artifacts compared with the single TE acquisition. (Adapted from Bolan, et al [8]).

上一篇 活体肝细胞凋亡代谢改变的质子磁共振频谱研究
下一篇 弥漫性轴索损伤的MRI研究进展
  
诚聘英才 | 广告合作 | 免责声明 | 版权声明
联系电话:010-67113815
京ICP备19028836号-2