分享:
分享到微信朋友圈
X
临床研究
探讨1H-MRS定量分析在脑实质区胶质瘤分级中的诊断价值
姜亮 刘文 肖朝勇 李宗鸿 胡军 殷信道

姜亮,刘文,肖朝勇,等.探讨1H-MRS定量分析在脑实质区胶质瘤分级中的诊断价值.磁共振成像, 2015, 6(1): 15-20. DOI:10.3969/j.issn.1674-8034.2015.01.004.


[摘要] 目的 定量分析氢质子磁共振波谱(1H-MRS)代谢产物在脑实质区胶质瘤分级诊断中的诊断价值。材料与方法 回顾性分析经病理证实的58例脑实质区胶质瘤的1H-MRS图像。按照世界卫生组织(WHO)2007标准,将病例分为两组:低级别胶质瘤组(Ⅰ~Ⅱ级,24例)和高级别胶质瘤组(Ⅲ~Ⅳ级,34例)。测量肿瘤实质区代谢物及对侧相应脑组织代谢物的含量并进行统计学分析。结果 (1)58例脑胶质瘤实质区的rCho、Cho/Cr较对侧正常区均不同程度升高,其余代谢物及比值呈不同程度降低。在两组胶质瘤中,肿瘤实质区rNAA、Cho/Cr、NAA/Cr、NAA/Cho值及Lac峰存在显著差异(P<0.05)。(2)肿瘤实质区rCho、Cho/Cr与肿瘤级别呈正相关,rNAA、NAA/Cr、NAA/Cho与肿瘤级别呈负相关,其中rNAA及NAA/Cho的相关性较高,相关系数r分别为-0.555、-0.616。(3)经ROC分析发现rNAA、NAA/Cr、NAA/Cho鉴别胶质瘤分级的曲线下面积大于0.5,肿瘤实质区rNAA鉴别胶质瘤分级的最佳截断值为0.20(敏感度为100%,特异度为61.8%);肿瘤实质区NAA/Cr鉴别胶质瘤分级的最佳截断值为0.35(敏感度为100%,特异度为55.9%);肿瘤实质区NAA/Cho鉴别胶质瘤分级的最佳截断值为0.19(敏感度为91.7%,特异度为67.6%)。结论 依据1H-MRS定量参数分析,肿瘤实质区NAA、NAA/Cr、NAA/Cho鉴别脑实质区胶质瘤分级具有很高的灵敏度和稍高特异度,有助于脑实质区胶质瘤的分级诊断,而肿瘤实质区rCho、Cho/Cr在脑实质区胶质瘤分级的诊断价值还有待进一步研究。
[Abstract] Objectives: To study the value of quantitative analysis of proton magnetic resonance spectroscopy (1H-MRS) metabolites in diagnosis of Brain Parenchyma glioma grading.Materials and Methods: The 1H-MRS metabolites of total 58 Brain Parenchyma gliomas confirmed by pathology were analyzed retrospectively. According to the 2007 World Health Organization (WHO) standards, all cases were divided into two groups including low-grade gliomas group (grade Ⅰ—Ⅱ, 24 cases) and high-grade gliomas group (grade Ⅲ—Ⅳ, 35 cases). Then concentrations of all metabolites in the tumor area and the contralateral brain tissue were analyzed statistically and compared.Results: (1) In total 59 glioma cases; Cho and Cho/Cr in tumor area were increased compared to that in the contralateral one while the other metabolites and metabolic ratio were decreased. There were significant differences in rNAA, Cho/Cr,NAA/Cr, NAA/Cho, and Lac between low-grade and high-grade glioma tumor(P<0.05). (2) rCho and Cho/Cr in the solid area of gliomas were positively related with pathological grading, rNAA, NAA/Cr, and NAA/Cho were negatively correlated with pathological grading. rNAA and NAA/Cho had higher correlation than the others, and the r correlation coefficient was 0.555 and 0.616 respectively. (3) The area under the curve (AUC) of rNAA, NAA/Cr and NAA/Cho for identifying glioma grading were greater than 0.5 by the ROC analysis. The cutoff point of rNAA for identifying glioma grading was set at 0.20 with 100% sensitivity and 61.8% specifity. The cutoff point of NAA/Cr for identifying glioma grading was set at 0.35 with 100% sensitivity and 55.9% specificity. The cutoff point of NAA/Cho for identifying glioma grading was set at 0.19 with 91.7% sensitivity and 67.6% specificity.Conclusions: According to quantitative analysis of 1H-MRS parameters, the measurements of NAA, NAA/Cr, and NAA/Cho which have high sensitivity and moderate specificity can provide additional information about the diagnosis of glioma grading, whereas the diagnosis value of rCho and Cho/Cr in glioma grading still needs further research.
[关键词] 神经胶质瘤;磁共振波谱学;代谢
[Keywords] Glioma;Magnetic resonance spectroscopy;Metabolism

姜亮 南京医科大学附属南京医院(南京市第一医院)医学影像科,南京 210006

刘文 南京医科大学附属脑科医院放射科,南京 210029

肖朝勇* 南京医科大学附属脑科医院放射科,南京 210029

李宗鸿 南京医科大学附属脑科医院放射科,南京 210029

胡军 南京医科大学附属脑科医院放射科,南京 210029

殷信道* 南京医科大学附属南京医院(南京市第一医院)医学影像科,南京 210006

通讯作者:殷信道,E-mail:y.163yy@163.com 肖朝勇,E-mail: 13851424310@163.com


收稿日期:2014-07-28
接受日期:2014-09-28
中图分类号:R445.2; R730.264 
文献标识码:A
DOI: 10.3969/j.issn.1674-8034.2015.01.004
姜亮,刘文,肖朝勇,等.探讨1H-MRS定量分析在脑实质区胶质瘤分级中的诊断价值.磁共振成像, 2015, 6(1): 15-20. DOI:10.3969/j.issn.1674-8034.2015.01.004.

       脑胶质瘤是神经系统最常见的恶性肿瘤,约占颅内肿瘤的40%左右,其发病有逐年上升趋势[1],其中脑实质区胶质瘤最常见,胶质瘤具有侵袭性生长、对放化疗不敏感等特性,是占首位死因的脑肿瘤[2,3]

       磁共振成像(magnetic resonance imaging,MRI)已成为评价胶质瘤的首选影像学方法,但MRI平扫和增强图像对胶质瘤良恶性鉴别存在局限性[4],准确率约55%~83.3%[5]。氢质子磁共振波谱(1H-magnetic resonance spectroscopy ,1H-MRS)是目前惟一能无创性探测活体正常脑组织和颅内肿瘤组织代谢变化的方法,从分子水平分析肿瘤,反映组织内部基本代谢变化,从而推断该组织的病理学变化,弥补了常规MRI检查的不足。笔者对58例经手术病理证实的脑实质区胶质瘤患者的1H-MRS资料进行分析,探讨1H-MRS定量参数在脑实质区胶质瘤分级中的诊断价值。

1 材料与方法

1.1 临床资料

       搜集2011年4月1日至2013年12月1日在南京医科大学附属脑科医院诊治并经病理证实的58例脑实质区胶质瘤患者的资料,所有患者术前均行MRI平扫加增强、1H-MRS检查。按世界卫生组织(World Health Organization,WHO)分级2007标准[6]Ⅰ级1例,Ⅱ级23例,Ⅲ级16例,Ⅳ级18例,男33例,女25例,平均年龄(49±14)岁。患者主要症状为头痛、恶性、呕吐或发作性四肢抽搐,意识丧失。

1.2 MRI和MRS检查方法

       采用Siemens 3.0 T超导磁共振机检查,常规MRI扫描及1H-MRS扫描均采用标准正交头线圈,MRI常规平扫参数:轴面T1WI (TR 250 ms,TE 2.8 ms)、T2WI(TR 3000 ms,TE 104 ms),层厚6 mm及横断面DWI (b值=1000 s/mm2)。增强扫描序列包括矢状面、横断面及冠状面T1,对比剂采用Gd-DTPA(马根维显):0.2 mmol/kg,经肘前静脉注射。1H-MRS扫描:采用二维多体素波谱成像(2D-MRSI),选择肿瘤最大层面作为MRS定位的层面,感兴趣区(ROI)根据病变部位、范围而定,既要尽可能涵盖肿瘤实质以及对侧正常脑组织,也要避开骨骼、脂肪及含气结构。采用点分辨波谱(2D-PRESS)序列扫描,TR 250 ms,TE 2.8 ms,FOV 120 mm×120 mm~140 mm×140 mm,激励次数为1次,成像时间为413 s。

1.3 图像及数据分析

       由2名高年资MRI医师使用仪器自带软件包Syngo对图像进行后处理。在波谱定位图上选择多个ROI ,ROI放置于肿瘤实性部分(取肿瘤强化程度最大的区域)和对侧正常脑组织区。在波谱图上观察代谢物的变化情况,波谱自带软件计算出代谢物N-乙酰天冬氨酸(NAA 2.02 ppm)、肌酸(Cr 3.02 ppm)、含胆碱代谢物(Cho 3.22 ppm)等物质的波峰下面积,计算代谢物相对值rCho、rCr、rNAA(肿瘤实质区各代谢物与健侧正常脑组织区相应值相比)及Cho/Cr、NAA/Cr、NAA/Cho比值,选取三组基线平稳的ROI并计算平均值作为最终各代谢物的相对值及比值。

1.4 统计分析

       所测数据用IBM SPSS19.0软件包进行统计分析,统计结果均用平均数±标准差表示,采用独立t检验比较低级别组和高级别组的rCho、rCr、rNAA、Cho/Cr、NAA/Cr、NAA/Cho;运用Mann-Whitney U检验比较两组Lac峰;各代谢物值与肿瘤级别的相关性分析采用Spearman非参数相关分析;P<0.05为差异有统计学意义。运用受试者操作曲线(receiver operating characteristic curve,ROC)分析代谢物鉴别胶质瘤分级的可行性,以曲线下面积大者(under the curve,AUC)为最佳预测值,按照约登指数(灵敏度+特异度-1)最大确定临界值及其敏感度和特异度。

2 结果

2.1 低、高级别组1H-MRS代谢物比较

       58例脑胶质瘤病灶的Cho、Cho/Cr值较对侧正常区均不同程度升高,其余代谢物及比值呈不同程度降低(图1图2)。经独立t检验分析,低级别组与高级别组胶质瘤肿瘤实质区rCho及rCr无统计学差异(P>0.05) ;rNAA、Cho/Cr、NAA/Cr及NAA/Cho在低、高级别组胶质瘤比较中均有统计学意义(P<0.05)(表1)。24例低级别胶质瘤中2例显示较低的Lac峰,34例高级别胶质瘤中11例出现Lac峰,两组Lac峰差异有统计学意义(P<0.05)。

图1  女,72岁,右颞顶叶少突-星形胶质细胞瘤(Ⅱ级)。A、B:肿瘤实质区,Cho峰显著升高,NAA峰和Cr峰明显降低,NAA/Cr=0.47、NAA/Cho=0.17、Cho/Cr=2.72。C、D:为对侧正常脑组织区MRS代谢物值,由此可得出rCho=2.66、rNAA=0.23、rCr=0.98
图2  男,61岁,左额叶间变型星形胶质细胞瘤(Ⅲ级)。A、B:肿瘤实质区,Cho峰升高,NAA峰和Cr峰降低,NAA/Cr=0.52、NAA/Cho=0.11、Cho/Cr=4.59。C、D:为对侧正常脑组织区MRS代谢物值,由此可得出rCho=1.40、rNAA=0.16、rCr=0.33
Fig. 1  72-year-old-woman, a histologically verified grade II oligodendroglioma in the right temporo-parietal lobe. A, B: The Cho peak significantly rise and the NAA ,Cr peak was declined in the tumor area. NAA/Cr=0.47, NAA/Cho=0.17, Cho/Cr=2.72. C, D: The MRS metablites values of the contralateral mirroring brain tissue, we can conclude that rCho=2.66, rNAA=0.23, rCr=0.98.
Fig. 2  61-year-old-man, a histologically verified grade Ⅲ anaplastic astrocytomas in the left frontal lobe. A, B: The Cho peak rise and the NAA, Cr peak was declined in the tumor area. NAA/Cr=0.52, NAA/Cho=0.11, Cho/Cr=4.59. C, D: The MRS metablites values of the contralateral mirroring brain tissue, we can conclude that rCho=1.40, rNAA=0.16, rCr=0.33.
表1  低、高级别组胶质瘤的肿瘤区1H-MRS代谢物比较(+s)
Tab. 1  Comparison of 1H-MRS metablites values between the high- and low-grade groups of gliomas (+s)

2.2 1H-MRS代谢物与肿瘤级别相关性分析

       Spearman非参数相关分析显示脑胶质瘤肿瘤实质区rCho、Cho/Cr与肿瘤级别呈正相关(r=0.29 ,P=0.027 ;r=0.468 ,P=0.000) ,rNAA、NAA/Cr、NAA/Cho与肿瘤级别呈负相关(r=-0.555,P=0.000;r=-0.418 ,P=0.001 ;r=-0.616 ,P=0.000)。

2.3 ROC曲线分析1H-MRS代谢物对胶质瘤分级的灵敏度和特异度

       经ROC分析发现肿瘤实质区rCho、Cho/Cr诊断胶质瘤低、高级别的曲线下面积小于0.5,无统计学意义。rNAA、NAA/Cr、NAA/Cho诊断胶质瘤低、高级别的曲线下面积分别为0.825、0.745、0.861(图3);以rNAA 0.20为阈值,鉴别低、高级别胶质瘤诊断敏感度为100%,特异度为61.8%;以NAA/Cr 0.35为阈值,鉴别低、高级别胶质瘤诊断敏感度为100%,特异度为55.9%;以NAA/Cho 0.19为阈值,鉴别低、高级别胶质瘤诊断敏感度为91.7%,特异度为67.6%。

图3  1H-MRS代谢物鉴别胶质瘤分级的ROC曲线
Fig. 3  ROC curves of 1H-MRS metablites in differentiating gliomas grading.

3 讨论

       脑胶质瘤高低级别的准确诊断对手术方案的制订和预后判断具有重要价值,高级别胶质瘤被误诊为低级别胶质瘤会导致治疗不足,相反则会导致过度治疗,这两者都会在一定程度上增加复发率或病死率。

       常规MRI主要依据典型的信号、形态、部位、强化程度及占位效应对胶质瘤作出诊断,而对胶质瘤低、高级别的鉴别存在一定局限性。1H-MRS主要利用化学位移的原理反映脑内主要代谢物的水平,能从细胞水平无创性的反映代谢产物的病理生理变化[7,8],从而提高肿瘤分级的准确度,指导临床手术的制订和预后判断[9]

       本研究为回顾性研究,58例患者的图像均无伪影,扫描模式均相同,MRS扫描均采用点分辨波谱(2D-PRESS)法,因2D-PRESS具有诸多优点,如信噪比较高,信号采集时间缩短,波谱分辨率高,基线稳定,对各种代谢物的定量更加可靠。另本研究在肿瘤实质区及对侧正常脑组织区放置多个ROI,并选取基线平稳的MRS取平均值,使本组数据可靠性更高。Cho峰的高低可作为肿瘤细胞增殖的指标,胆碱的增高可提示细胞快速增殖[10]

       本研究结果显示高级别组rCho和Cho/Cr值高于低级别组,且Cho/Cr在低、高级别组胶质瘤中具有统计学差异,可见恶性程度较高的胶质瘤rCho和Cho/Cr值一般较高。NAA主要存在于成熟的神经元内,其含量降低代表神经元功能受损[11]。脑胶质瘤是由异常增生的胶质类细胞组成,在生长过程中,肿瘤细胞浸润正常神经元,本研究结果显示高级别组rNAA、NAA/Cr及NAA/Cho值低于低级别组,差异有统计学意义,由此可见高级别胶质瘤细胞增殖快、神经元破坏更明显。笔者发现,在高、低级别组胶质瘤中均会出现异常的Lac峰,且大部分出现在高级别组,Lac峰的出现提示正常细胞的有氧呼吸被抑制,表明脑组织缺血、缺氧,甚至脑死亡[12],说明高级别胶质瘤往往存在坏死,Lac峰的出现对胶质瘤的分级具有一定的价值,这与其他学者研究结果一致[13]

       随着胶质瘤级别升高,Cho呈升高趋势,NAA呈下降趋势,相关性分析显示,脑胶质瘤肿瘤实质区的rCho、rNAA、Cho/Cr、NAA/Cr及NAA/Cho值与肿瘤级别存在相关性,但rCho与肿瘤级别相关性较低,仅为0.29,而rNAA及NAA/Cho与肿瘤级别的相关性较高,rNAA与肿瘤级别的相关性较高的原因可能为高级别胶质瘤神经元破坏程度远远大于低级别胶质瘤,正常组织被肿瘤细胞浸润的程度较大,但rNAA值易受个体差异影响,而NAA/Cho值对胶质瘤的分级准确性更高,这是因为NAA反映的是正常组织被肿瘤细胞浸润、代替程度,Cho反映肿瘤细胞分裂增殖情况,而胶质瘤的病理分级是由细胞核的异型性(染色体粗糙,核多形性和多核)、核分裂活跃、细胞密度增高、血管内皮细胞增生和坏死决定[14,15,16]

       笔者通过ROC曲线对各代谢物区分高低级别胶质瘤的效能进行评估,显示肿瘤实质区rCho、Cho/Cr的AUC小于0.5,不具有诊断价值,而肿瘤实质区rNAA、NAA/Cr、NAA/Cho的AUC均大于0.5,其中NAA/Cho的AUC最大(0.861),可见肿瘤实质区NAA/Cho的诊断价值优于rNAA、NAA/Cr。通过该曲线可得到区分低级别、高级别肿瘤较好的界值点,理想的界值点应该具有较高诊断率及较低误诊率。本研究结果显示,肿瘤实质区rNAA、NAA/Cr的灵敏度均达到100%,但特异度偏低,分别为61.8%、55.9%;肿瘤实质区NAA/Cho的灵敏度较rNAA、NAA/Cr稍低,但特异度高于rNAA、NAA/Cr,这与其他学者研究的结果存在些许差异[16,17,18],可能原因为样本量不同,包括高级别、低级别样本量;1H-MRS扫描方法不同,包括MR磁场强度、采集参数、体素大小和放置位置[19,20]。研究中,笔者发现肿瘤实质区rCho、Cho/Cr值与肿瘤级别存在相关性,但相关性较低,且ROC曲线分析其AUC小于0.5,不具有诊断价值,这与其他学者结果不一致[21],可能原因为病理类型及样本量不同,本研究所选病例均为脑实质区胶质瘤,病理类型不包括髓母细胞瘤、室管膜瘤、脑膜瘤等。可见Cho在胶质瘤低、高级别分级中的诊断价值有待进一步研究。

       本研究还存在一些不足之处,如仅将胶质瘤分为低、高级别,尚未达到WHO分级标准(Ⅰ~Ⅳ);未分析瘤周水肿区各代谢物及比值的统计学意义;未统计Lac峰值。有待进行更深入、细致的研究。

       综上所述,1H-MRS定量参数的测量有助于脑实质区胶质瘤分级,尤其是肿瘤实质区NAA、NAA/Cr、NAA/Cho鉴别胶质瘤分级具有很高的灵敏度和稍高特异度,有助于胶质瘤的分级诊断,而肿瘤实质区rCho、Cho/Cr在胶质瘤分级的诊断价值还有待进一步研究。

[1]
Dolecek TA, Propp JM, Stroup NE, et al. CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2005-2009. Neuro Oncol, 2012, 15(5): 646-647.
[2]
Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med, 2005, 352(10): 987-996.
[3]
Negendank W, Sauter R. Intratumoral lipids in 1H MRS in vivo in brain tumors: experience of the Siemens cooperative clinical trial. Anticancer Res, 1996, 16(3B): 1533-1538.
[4]
Al-Okaili RN, Krejza J, Woo JH, et al. Intraaxial brain masses: MR imaging-based diagnostic strategy: initial experience. Radiology, 2007, 243(2): 539-550.
[5]
Goebell E, Fiehler J, Ding XQ, et al. Disarrangement of fiber tracts and decline of neuronal density correlate in glioma patients-a combined diffusion tensor imaging and 1H-MR spectroscopy study. AJNR Am J Neuroradiol, 2006, 27(7): 1426-1431.
[6]
Louis DN, Ohgaki H, Wiestler OD, et al. The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol, 2007,114(2): 97-109.
[7]
Cousins JP. Clinical MR spectroscopy: fundamentals, current applications, and future potential. AJR Am J Roentgenol, 1995, 164(6): 1337-1347.
[8]
Huang L, Huang SQ, Yuan XP, et al. The correlation of 1H-MRS, ADC value and proliferative activity Ki-67 of glioma in peritumoral edema region and parenchyma region. J Pract Radiol, 2013, 29(6): 875-879.
黄莉,黄穗乔,袁小平,等.胶质瘤瘤体及瘤周1H-MRS、ADC值与Ki-67的相关性研究.实用放射学杂志, 2013, 29(6): 875-879.
[9]
Bulik M, Jancalek R, Vanicek J, et al. Potential of MR spectroscopy for assessment of glioma grading. Clin Neurol Neurosurg, 2013,115(2): 146-153.
[10]
Wu YQ, Lin Q, lan YH, et al. Glioblastoma multi-mode MRI manifestations and pathological histology foundation. Chin J Magn Reson Imaging, 2013, 4(3): 196-200.
吴裕强,林祺,兰玉华,等.胶质母细胞瘤多模式MRI表现及其病理组织学基础.磁共振成像, 2013, 4(3): 196-200.
[11]
Yin H, Zhang J, Gao YG, et al. Application of magnetic resonance spectroscopy in diagnosis of intracranial lesions. J Pract Radiol, 2007, 23(6): 721-723.
印弘,张艰,高元桂,等. MRS在脑占位病变诊断中的应用.实用放射学杂志, 2007, 23(6): 721-723.
[12]
He D, Yang L, Zhu QF, et al. The value of 1H-MRS in cerebral glioma. J Chin Clin Med Imaging, 2011, 22(2):77-82.
贺丹,杨丽,朱青峰,等. 1H-MRS在颅内胶质瘤分级中的应用价值.中国临床医学影像学杂志, 2011, 22(2): 77-82.
[13]
Lin Q, Zhang Q, Chen DP, et al.Correlation investigate between grade of intratumoral susceptibility signals and relative quantitative of 1H-MRS in patients with brain astrocytic tumours. Chin J Magn Reson Imaging, 2012, 3(2): 174-178.
林祺,张强,陈东平,等.脑星形细胞瘤磁敏感效应级别与1H-MR波谱相对定量的相关性探讨.磁共振成像, 2012, 3(2): 174-178.
[14]
Oshiro S, Tsugu H, Komatsu F, et al. Quantitative assessment of gliomas by proton magnetic resonance spectroscopy. Anticancer Res, 2007, 27(6A): 3757-3763.
[15]
Likavcanová K, Dobrota D, Liptaj T, et al. In vitro study of astrocytic tumour metabolism by proton magnetic resonance spectroscopy. Gen Physiol Biophys, 2005, 24(3): 327-335.
[16]
Zeng Q, Liu H, Zhang K, et al. Noninvasive evaluation of cerebral glioma grade by using multivoxel 3D proton MR spectroscopy. Magn Reson Imaging, 2011, 29(1): 25-31.
[17]
Stadlbauer A, Gruber S, Nimsky C, et al. Preoperative grading of gliomas by using metabolite quantification with high-spatial-resolution proton MR spectroscopic imaging. Radiology, 2006, 238(3): 958-969.
[18]
Peng HJ, Yao QW, Zhao B, et al. The value of single-voxel 1H-MRS in differentiating of benign and malignant gliomas. J Med Imaging, 2010, 20(12): 1761-1763.
彭洪娟,姚其卫,赵斌,等.单体素1H-MRS在良恶性胶质瘤鉴别诊断中的价值.医学影像学杂志, 2010, 20(12): 1761-1763.
[19]
Soares DP, Law M. Magnetic resonance spectroscopy of the brain: review of metabolites and clinical applications. Clin Radiol, 2009, 64(1): 12-21.
[20]
Li BS, Regal J, Gonen O. SNR versus resolution in 3D 1H-MRS of the human brain at high magnetic fields. Magn Reson Med, 2001, 46(6): 1049-1053.
[21]
Wang Y, Qi JP, Cheng SR, et al. Ideal Cho/Cr threshold of 1H-MRS in Evaluation of the pathological grade in gliomas. Mil Med J S Chin, 2014, 28(1):15-18.
王鹰,漆剑频,程少容,等.氢质子波谱分析Cho/Cr值评估胶质瘤病理分级的理想阈值研究.华南国防医学杂志, 2014, 28(1): 15-18.

上一篇 首发抑郁症脑局部一致性静息态MRI对比研究
下一篇 慢性梗阻性肺部病变患者脑皮质MRI形态学初步研究
  
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