【作者】 冯霞；

【导师】 邱士军；

【作者基本信息】 南方医科大学 ， 影像医学与核医学， 2013， 博士

【摘要】 第一部分DTI定量分析鼻咽癌放疗后常规MRI显示阴性的脑白质扩散张量改变的应用研究目的：通过测量鼻咽癌放疗前及放疗后不同阶段常规MRI显示阴性的颞叶脑白质的FA值、λ‖及λ⊥,并进行纵向对比分析,探讨放疗对颞叶白质水分子扩散的影响,以及扩散特征随时间变化的规律。评估DTI显示和追踪放疗后常规MRI显示阴性的颞叶脑白质微观病理改变的可行性与临床应用价值。材料和方法：对64例鼻咽癌患者(男46例,女18例,年龄18-65岁,平均年龄44岁)进行磁共振扩散张量成像(DTI)检查,其中放疗前16例,放疗后起48例。放疗前患者归入对照组：放疗后患者依据国内外较常采用的放射治疗后分期方法分为3组：组1(GoupⅠ),放疗后0-6个月,即放射性脑损伤的急性期与早期迟发性反应期；组2(GoupⅡ),放疗后6-12个月,即放射性脑损伤的晚期迟发性反应期；组3(GoupⅢ),放疗后6-12个月,仍为放射性脑损伤的晚期迟发性反应期,每组16例。放疗前后各组年龄、性别、受教育程度、吸咽及饮酒量匹配。所有患者均经病理检查证实；放疗后患者均为首次接受放射；放射治疗方案为头颈联合野三维适形放疗(总剂量／分割剂量／分割次数,66-72Gy/2Gy,分割次数为33～36),总放射治疗时间为43～51天,所有患者在放疗中接受同部化疗,主要的化疗药物为顺铂、环磷酰胺等：所有患者未发现鼻咽癌颅内侵犯,无颅内原发性肿瘤或转移瘤,无颅内血管性病变,无严重的系统性疾病如心脏病、糖尿病、高血压等,放疗前后均未出现明显的神经系统症状,常规MRI扫描脑组织表现无明显异常。使用飞利浦1.5T双梯度磁共振扫描仪进行DTI数据采集,接收核磁共振信号使用16通道神经血管线圈。DTI数据采集采用单次激发的自旋回波平面回波序列,平行于大脑前后联合得到全脑轴位弥散加权成像。扩散敏感梯度方向为33个,扩散敏感系数b=0和b=800s/mm2,重复时间(TR)=10793ms；回波时间(TE).62ms；翻转角(flip angle)=90°；矩阵(matrix size)=128x128；视野(FOV)=21×21cm；层间隔=0mm,激励次数(NEX)=I；共50层。DTI数据分析使用DTIStudio软件,原始的DICOM数据经MRIcro软件转化为四维NIFTI格式数据后,输入DTIStudio进行头动校正及涡流较正,去除头部范围以外的噪声部分,再由软件中特定函式计算扩散张量,并输出扩散张图。选择感兴趣的扩散张量图(FA、λ‖及λ⊥),在大脑脚层面双侧颞叶前部脑白质内各取一个面积相等的感兴趣区(ROI),计算双侧颞叶两个ROI的平均值作为最后值。四组被试间年龄、受教育程度、吸烟数和饮酒量的差异采用单因素方差分析的方法进行组间比较。四组间性别差异采用卡方检验。四组间颞叶白质平均FA值、λ‖值以及λ⊥值采用单因素方差分析比较组间差异后,再进行两两组间比较。P<0.05认为有统计学意义。以上所有分析均通过SPSS13.0完成。结果：1.FA值四组间FA值差异有统计学意义。放疗后0-6个月FA值下降最为明显,与放疗前相比差异有显著性；放疗后6-12个月,FA值比放疗后0-6个月明显升高,但仍低于放疗前水平,差异均有显著性；放疗后>12个月,FA值比放疗后6～12个月有升高趋势,但仍显著低于放疗前水平。2.λ‖四组间λ‖值差异有统计学意义。放疗后0-6个月λ‖值下降最为明显,与放疗前相比差异有显著性；放疗后6-12个月,λ‖值比放疗后0-6个月表现出升高趋势,但仍显著低于放疗前水平；放疗后>12个月,λ‖值比放疗后6-12个月进一步升高,但仍显著低于放疗前水平。3.λ⊥四组间λ⊥值差异有统计学意义。放疗后0-6个月λ⊥值升高最为明显,与放疗前相比差异有显著性；放疗后6-12个月,λ⊥值比放疗后0-6个月有所回落,但仍然高于放疗前水平；放疗后>12个月,λ⊥值比放疗后6-12个月进一步回落,但仍略高于放疗前水平。结论：1.放疗后常规MRI显示阴性的颞叶白质λ‖明显降低而λ⊥明显升高,其病理学基础分别为轴突损伤和脱髓鞘等微观结构组织学改变；FA值显著减小,其病理基础除轴突损伤和脱髓鞘等神经纤维等微观结构损伤以外,可能还因血管损伤、神经炎症及继发的间质水肿所致。2.放疗后颞叶白质水分子扩散改变呈动态过程,急性期与早期迟发性反应期(放疗后0-6个月)λ‖、λ⊥及FA值改变最明显,而随着放疗后时间的延长,放疗后1年以后双侧颞叶λ‖、λ⊥及FA值均恢复到接近放疗前水平,与轴突再生、髓鞘磷脂化、血管损伤修复、炎症吸收及水肿消退等修复过程有关。总结放疗后颞叶白质的扩散特征改变的规律,为放疗后不同阶段的改变提供参考依据,有利于监测放射性脑损伤的程度及修复过程,有助于及时发现严重的病变及病变的突然进展,及时治疗,把放射性脑损伤的危害减小到最低程度。3.DTI以探测活体组织水分子扩散运动为基础,可以敏感的探测脑组织受照射后潜伏期内λ‖、λ⊥及FA值的变化情况,进而反应放射性脑损伤潜伏期微观病理改变。以DTI为手段对照射后的脑组织进行动态监测,及时发现病变的突然进展将有助于放射性脑病的早期诊断。第二部分DSC-PWI定量分析鼻咽癌放疗后常规MRI显示阴性的脑白质灌注参数改变的应用研究目的：通过测量鼻咽癌放疗前及放疗后不同阶段常规MRI显示阴性的颞叶脑白质rCBF值,并进行纵向对比分析,探讨放疗对颞叶白质局部相对血流量的影响,以及相对血流量随时间变化的规律。评估DSC-PWI显示和追踪放疗后常规MRI显示阴性的脑白质微观病理改变的可行性与临床应用价值。材料和方法：对64例鼻咽癌患者(男46例,女18例,年龄18-65岁,平均年龄44岁)进行动态敏感对比增强灌注加权成像(DSC-PWI)检查,其中放疗前16例,放疗后起48例。放疗前患者归入对照组。放疗后患者依据国内外较常采用的放射治疗后分期方法分为3组：组1(GoupⅠ),放疗后0～6个月,即放射性脑损伤的急性期与早期迟发性反应期；组2(GoupⅡ),放疗后6～12个月,即放射性脑损伤的晚期迟发性反应期；组3(GoupⅢ),放疗后6～12个月,仍为放射性脑损伤的晚期迟发性反应期；每组16例。放疗前后各组年龄、性别、受教育程度、吸咽及饮酒量匹配。所有患者均经病理检查证实；放疗后患者均为首次接受放射；放射治疗方案为头颈联合野三维适形放疗(总剂量/分割剂量/分割次数,66～72Gy/2Gy,分割次数为33-36),总放射治疗时间为43-51天,所有患者在放疗中接受同步化疗,主要的化疗药物为顺铂、环磷酰胺等；所有患者未发现鼻咽癌颅内侵犯,无颅内原发性肿瘤或转移瘤,无颅内血管性病变,无严重的系统性疾病如心脏病、糖尿病、高血压等,放疗前后均未出现明显的神经系统症状,常规MRI扫描脑组织表现未见明显异常。使用飞利浦1.5T双梯度磁共振扫描仪进行DSC-PWI数据采集。接收核磁共振信号使用16通道神经血管线圈。DSC-PWI数据采集采用单次激发自旋回波平面回波序列(SE-EPI),平行于大脑前后联合得到全脑轴位灌注加权成像。磁共振扫描具体参数如下：重复时间(TR)=1000ms；回波时间(TE)=30ms:视野(FOV)=202×202mm2；矩阵(matrix size)=128×128；voxel size=1.9×1.9×1.9mm3；层厚=5mm；层间隔=0.30mm；45个时相,激励次数(NEX)=1次。扫描时间为89s,产生495幅颞叶灌注图像。对比剂为亚喷酸葡铵(Gd-DTPA),剂量按0.2mmol/kg体重计算,注射速率为5mL/s,由21号套管针经肘前静脉注射,随后以相同速率注射20mL生理盐水冲刷,注射过程由高压注射器控制。采用单因素方差分析的方法对四组被试年龄、受教育程度、当前每天吸烟数量、当前饮酒量分别进行组间分析；四组间性别差异采用卡方检验。四组间颞叶白质平均rCBF(相对血流量)采用单因素方差分析比较组间差异后,再进行两两组间比较。P<0.05认为有统计学意义。以上所有分析均通过SPSS13.0完成。结果：四组间rCBF值差异有统计学意义。放疗后0～6个月rCBF值下降最为明显,与放疗前相比差异有显著性；放疗后6～12个月,rCBF值比放疗后0～6个月明显升高,但仍显著低于放疗前水平；放疗后>12个月,rCBF值比放疗后6～12个月进一步升高,但仍显著低于放疗前水平。结论：1.放疗后常规MRI显示阴性的颞叶白质rCBF明显降低,其病理生理学基础为血管内皮损伤、基底膜损伤及神经炎症等,导致血脑屏障破坏及血管壁通透性增高,影响了血管的结构及功能状况。血管损伤及其所致的血流量减低对于全身需氧量最高的脑组织必然是重要的损伤因素,因而在前面的研究中我们看到了轴突损伤和脱髓鞘等神经纤维的微观结构损伤。2.放疗后颞叶白质血流动力学改变呈动态过程,急性期与早期迟发性反应期(放疗后0～6个月)rCBF值改变最明显,而随着放疗后时间的延长,rCBF值逐渐恢复,表明放射性血管损伤存在修复机制。随着血管损伤不断修复,神经组织供血供氧得到改善,也必然促进轴突再生、髓鞘磷脂化等神经组织的修复过程。结合前面的研究,我们也看到血管损伤修复的同时伴随着轴突再生、髓鞘磷脂化等神经组织的修复现象。总结放疗后颞叶白质血流动力学改变的规律,为放疗后不同阶段的改变提供参考依据,有利于监测放射性脑血管损伤的程度及修复过程,从而在神经组织坏死前早期发现血管损伤的突然进展,及时治疗,把放射性脑损伤的危害减小到最低程度。3. DSC-PWI以探测活体组织血流动力学改变为基础,可以探测脑组织受照射后潜伏期内rCBF值的变化情况,进而反映放射性脑损伤潜伏期血管的微观病理改变,有利于放射性脑损伤发病机制的探索；同时,以DSC-PWI为手段对照射后的脑组织进行动态监测,及时发现血管病变的突然进展将有助于放射性脑病的早期诊断。第三部分脑结构成像分析鼻咽癌放疗后常规MRI显示阴性的脑灰质体积改变的应用研究目的：将鼻咽癌放疗后急性期与早期迟发性反应期常规MRI显示阴性的脑灰质体积与放疗前组比较,探讨放疗对脑灰质的影响,评估高分辨磁共振检查技术和DARTEL-VBM的分析方法测量常规MRI显示阴性的脑灰质结构改变的效果,为放射性脑损伤潜伏期内出现神经认知功能异常寻找答案。材料和方法：15例鼻咽癌患者(男12例,女3例,年龄34-54岁,平均年龄44岁)于治疗前接受首次结构磁共振成像检查,放疗结束后1-3月期间(即放射性脑损伤的急性期与早期迟发性反应期)接受第二次结构磁共振成像检查。所有患者均经病理检查证实,所有患者放射治疗方案接近,采用头颈联合野进行三维适形放疗,鼻咽部放射总剂量为68-70Gy,分割剂量为2Gy,分割次数为34-35次,总放射治疗时间为41～51天,所有患者在放疗中接受同步化疗,主要化疗药物为顺铂、环磷酰胺等；所有患者未发现鼻咽癌颅内侵犯,无颅内原发性肿瘤或转移瘤,无颅内血管性病变,无严重的系统性疾病如心脏病、糖尿病、高血压等,放疗前后均未出现明显的神经系统症状,常规MRI扫描脑组织表现肉眼可见的异常。使用飞利浦1.5T双梯度磁共振扫描仪进行数据采集,使用16通道神经血管线圈接收核磁共振信号。高分辨T1数据采集采用3D快速场回波序列(FFE),平行于正中矢状面得到全脑图像。具体参数如下：重复时间(TR)=25ms;回波时间(TE)=4.1ms;翻转角(flip angle)=30°;矩阵(matrix size)=231x231;视野(FOV)=230×230cm;层厚=1mm；层间隔=0mm,扫描层数在150-160层之间(因被试头颅的大小而异),描时间为6分35秒～7分02秒。图像预处理使用以SPM为平台的DARTEL工具箱,具体步骤如下：(1)手动调整所有被试的的原始结构图像；(2)分割(segment),对原始结构图像进行分割,并获得*seg.sn.mat文件,将其导入DARTEL内,以生成每位个体的灰质图像；(3)在DARTEL内制作被试的灰质模板(create template);(4)标准化(normalization),使用生成的灰质模板对所有被试者的灰质图像进行标准化；(5)调整(modulation):即对灰质密度图像进行Jacobian参数校正,以保证配准后的图像灰度值可以真实反映原图像的灰质结构体积。(6)平滑(smoothing):采用8mm的半高全宽(full width at half maximum, FWHM)的高斯核对分割后的灰质图像进行空间平滑,确保图像具有随机高斯场的性质,这样不仅满足了SPM的统计假设,又提高了信噪比。(7)空间转化：即把经过平滑处理后的数据转化为MNI空间。MRI图像预处理完成后,再应用SPM8软件进行后续分析。统计学分析以一般线性模型和随机场理论为基础,采用逐像素单边检验的方法,得到每个像素的t值,以P<0.05为检验水准,采用FDR校正,以200mmm3的体积阈值构成统计参数图,该图即表示鼻咽癌放疗后患者灰质体积改变的脑区,并以伪彩显不。结果：与放疗前相比,鼻咽癌患者放疗后右侧海马体积增大(P<0.05,FDR校正,体积阈值为100mm3)。没有显著容积减低区域。结论：1.放疗后右侧海马体积增大,可能是颞叶白质损伤后出现的功能代偿性肥大,也可能是由于放射线致该区脑组织损伤及其应对损伤出现的炎症反应所致,根据以往的动物实验研究,我们认为后前可能性大。2.放疗对脑灰质也有影响,而通过高分辨T1成像及基于VBM-DARTEL的分析方法,可以敏感的探测放疗后灰质的改变,是在活体开展放射性脑损伤研究的有效手段。更多还原

【Abstract】 Part one:The microstructure characteristics of temporal lobes white mater after radiation therapy for nasopharyngeal carcinomaObjective:To detect the microstructure characteristics of temporal lobes white mater after radiation therapy (RT) for nasopharyngeal carcinoma (NPC) using diffusion tensor imaging.Materials and methods:The study included64subjects (46male,18female, aged18-65years mean:44years), who had or had not undergone radiotherapy for nasopharyngeal carcinoma. All diagnoses were confirmed by histopathology. Of the64subjects,16were pre-treatment patients; the remaining48were post-radiation patients. The time from radiotherapy to imaging ranged from several days to4years. Three-dimensional conformal radiation therapy was used in the protocol for nasopharynx and neck radiotherapy. Nasopharynx radiotherapy involved a total dose of66-72Gy divided into33to36fractions, and was performed over a course of43to51days. Before patients underwent MRI examination, we confirmed that there were no signs of intracranial invasion of nasopharyngeal carcinoma or of intracranial tumor. Patients with white matter degeneration, vascular lesions, high blood pressure, diabetes, or heart disease were excluded from the study, as these lesions or diseases affect the DTI data measurements.The pre-radiation subjects constituted the control group. The post-radiation patients were divided into three groups according to the domestic and international staging systems for classifying radiation-induce brain injury. Group I included patients who had undergone radiation therapy less than6months prior, representing an acute response in the early period after radiation therapy. Post-radiation patients in the late delayed reaction period were divided into2groups:Group II included cases who had undergone radiotherapy6-12months prior while Group III included those who had undergone radiotherapy more than12months prior. Each group consisted of16subjects. There were no statistically significant differences between the groups with respect to age, gender, duration of education, cigarette smoking or alcohol consumption.MRI was performed using a Philips Achieva1.5-T Nova Dual MR scanner. DTI used a sixteen-channel head coil, obtaining DTI images. Diffusion imaging data were acquired using a32diffusion gradient direction scheme (b=800s/mm2) along32non-collinear directions plus b=0reference images acquired using a single-shot spin-echo planar sequence parallel to the AC-PC line to collect diffusion weighted images. Other imaging parameters were:TR=10793ms, TE=62ms, field of view=230×230mm2, matrix=128×128, slice thickness=2mm, with no slice gap, voxel size=2x2x2mm3.DTI post-processing used DTIStudio. After converting the raw perfusion data to the NIFTI format using MRIcro software, DTIStudio was applied to correct for head movement and eddy current. A fractional anisotropy (FA) map and eigenvalue images were generated based on the corrected DTI images. The FA value λ‖and λ⊥were calculated for each voxel.To find the layer of cerebral peduncle, one ROI were then marked on unilateral temporal lobe white matter in each subject. These ROIs were of equal area. The average of these values in the temporal lobes white matter was accepted as the final value.Statistical analysis was performed using SPSS version13.0software. One-way analysis of variance and two-sample t-tests were used to assess the differences in age, duration of education, cigarette smoking, and alcohol consumption across the groups. Pearson’sχtest was used to see the gender differences between the groups. All DTI data were examined for homogeneity of variance using one-way analysis of variance (ANOVA). The statistical differences between the groups were further examined using the Bonferroni method, with p<0.05accepted as statistically significant.Results:FA:Analysis of variance showed statistically significant differences between the groups with respect to FA value. Further group-wise comparisons showed that the FA value decreased significantly in the three post-radiation groups compared to the pre-radiation group; the FA value in Group Ⅱ increased significantly compared to Group Ⅰ; the FA value in Group Ⅲ was higher than that in Group Ⅱ, though with no significant difference, while it was significantly higher than in Group Ⅰλ‖:Analysis of variance showed statistically significant differences between the groups with respect to λ‖eigenvalue. Further group-wise comparisons showed that λ‖decreased significantly in the three post-radiation groups compared to the pre-radiation group; λ‖in Group Ⅲ was higher than that in Group Ⅱ, though with no significant difference, while it was significantly higher than in Group Ⅰ λ⊥:Analysis of variance showed statistically significant differences between the groups with respect to λ⊥. Further group-wise comparisons showed that λ⊥increased significantly in Group I compared to the pre-radiation group.Conclusion:The results of the present study illustrated microstructural changes occur in brain tissue that appears normal on routine MR imaging in patients with NPC after radiotherapy. The microscopic pathological changes in the early after irradiation of the brain improves some time after the cessation of radiotherapy. Radiation-induced brain injury of the temporal lobe exist for a considerable period of time, so making long-term clinical follow-up using DTI for dynamic monitoring of radioactive encephalopathy is of great clinical significance for radiotherapy patients who suffer from NPC. Part two:The hemodynamic characteristics of temporal lobes white mater after radiation therapy for nasopharyngeal carcinomaObjective:To detect the hemodynamic characteristics of temporal lobes white mater after radiation therapy (RT) for nasopharyngeal carcinoma (NPC) using dynamic susceptibility contrast-enhanced perfusion-weighted imaging.Materials and methods:The study included64subjects (46male,18female, aged18-65years mean:44years), who had or had not undergone radiotherapy for nasopharyngeal carcinoma. All diagnoses were confirmed by histopathology. Of the64subjects,16were pre-treatment patients; the remaining48were post-radiation patients. The time from radiotherapy to imaging ranged from several days to4years. Three-dimensional conformal radiation therapy was used in the protocol for nasopharynx and neck radiotherapy. Nasopharynx radiotherapy involved a total dose of66-72Gy divided into33to36fractions, and was performed over a course of43to51days. Before patients underwent MRI examination, we confirmed that there were no signs of intracranial invasion of nasopharyngeal carcinoma or of intracranial tumor. Patients with white matter degeneration, vascular lesions, high blood pressure, diabetes, or heart disease were excluded from the study, as these lesions or diseases affect the DSC-PWI data measurements.The pre-radiation subjects constituted the control group. The post-radiation patients were divided into three groups according to the domestic and international staging systems for classifying radiation-induce brain injury. Group Ⅰ included patients who had undergone radiation therapy less than6months prior, representing an acute response in the early period after radiation therapy. Post-radiation patients in the late delayed reaction period were divided into2groups:Group Ⅱ included cases who had undergone radiotherapy6-12months prior while Group Ⅲ included those who had undergone radiotherapy more than12months prior. Each group consisted of16subjects. There were no statistically significant differences between the groups with respect to age, gender, duration of education, cigarette smoking or alcohol consumption.MRI was performed using a Philips Achieva1.5-T Nova Dual MR scanner. Perfusion imaging used a sixteen-channel head coil, obtaining PWI images. Perfusion imaging of the temporal lobe was performed using a single-shot spin-echo version of an echo planar (SE-EPI) pulse sequence, parallel to the AC-PC line, with the following imaging parameters:TR=1000ms, TE=30ms, field of view=202×202mm2, matrix=115×116, voxel size=1.9×1.9×1.9mm3, slice thickness=5mm, slice gap=0.30mm, slice=11,45phases. A total of495images was acquired after an intravenous injection of0.2mmol/kg body weight of Gd-GTPA, with a5mL/s injection rate followed by a saline flush of20mL with the same injection rate using an automatic power injector. The acquisition time was89s.The Extended MR workspace was used for post-processing of PWI images. Software developed in-house (Neuro Perfusion software) was used to open the PWI scan data in the Extended MR workspace; the T2W imaging position image and the rCBF map were received automatically. Next, one ROI were drawn on unilateral temporal lobe in the layer of cerebral peduncle in each subject. These ROIs were of equal area. The average of the ROIs’ parameters’ values in the bilateral temporal lobe white matter was accepted as the final rCBF value.Statistical analysis was performed using SPSS version13.0software. One-way analysis of variance and two-sample t-tests were used to assess the differences in age, duration of education, cigarette smoking, and alcohol consumption across the groups. Pearson’s x test was used to see the gender differences between the groups. All DSC-PWI data were examined for homogeneity of variance using one-way analysis of variance (ANOVA). The statistical differences between the groups were further examined using the Bonferroni method, with p<0.05accepted as statistically significant.Results:The differences between the groups with respect to rCBF were statistically significant. Further group-wise comparisons showed that the rCBF in each post-radiation group decreased significantly compared to the pre-radiation group. The decrease was greatest in Group Ⅰ (0-6months after radiotherapy); the rCBF in Groups Ⅱ and Ⅲ was higher than that in Group Ⅰ but without significance; the rCBF was higher in Group Ⅲ than in Group Ⅱ but without significance. Conclusions:The results of the present study illustrated circulatory changes occur in brain tissue that appears normal on routine MR imaging in patients with NPC after radiotherapy. Vascular damage and the resulting decrease in blood flow may play a significant role in neural tissue injury in the early stages after radiotherapy. In later stages, improvement of the vascular damage may contribute to the recovery of radiation-induced brain injury. Vascular damage may play a significant role in radiation-induced brain injury. Radiation-induced brain injury of the temporal lobe exist for a considerable period of time, so making long-term clinical follow-up using PWI for dynamic monitoring of radioactive encephalopathy is of great clinical significance for radiotherapy patients who suffer from NPC. Part three:To explore the Gary Matter volume abnormalities after radiation therapy for nasopharyngeal carcinomaObjective:To explore the Gary Matter (GM) volume abnormalities in after radiation therapy (RT) for nasopharyngeal carcinoma by using high resolution MRI and DARTEL-VBM.Materials and methods:This study was carried out in8patients (12male,3female; age range,34-54years; mean age,44years) with a histopathologically established diagnosis of NPC, who received fractionated radiation therapies with3D conformal techniques (total doses/daily fraction/exposures/,68-70GY/2.0GY/34-35times). Each patient received MRI examinations before and1-2months after completion of the radiotherapy. No patient exhibited distinct neurological symptoms or showed signs of intracranial invasion or intracranial tumors before imaging. Patients were excluded from the study who presented with white matter degeneration, vascular lesions, hypertension, heart disease or diabetes.MRI was performed using a Philips Achieva1.5-T Nova Dual MR scanner. Perfusion imaging used a sixteen-channel head coil, obtaining3D-T1images.3D-T1imaging was performed using a three-dimensional fast field echo (FFE) pulse sequence, parallel to the AC-PC line, with the following imaging parameters:TR=4.1ms, TE=25ms, field of view=230×230mm2, matrix=231×231, slice thickness=1mm, slice gap=0mm, slice=1mm.The DARTEL tool, a extention tool of SPM, was used to perform anatomical images analysis. the main procedures include:(1) the naive image of all the subjects were segmented to obtain the*seg.sn.mat profile, which was then imported to DARTEL for calculateing and getting GM image for each participant.(2) create GM template using DARTEL.(3) use the GM template to normalize each participant’s GM with.(4) modulate the resulting GM images with a Jacobian determinant, then the voxel’s values indicate the absolute volume of the local GM.(5) use a12-mm full width at half maximum (FWHM) isotropic Gaussian kernel to smooth resulting images.(6) the GM was transformed to the MNI coordinate for SPM analysis. And then these preprocessed GM data was analysised by using SPM8through GLM and random field theory. The differences between two group was determined by a voxel based comparison. a cluster-level threshold of p<.05were set to produce statistical maps, which was corrected for multiple comparisons using false discovery rate (FDR).Results: The right hippocampus volume was significantly increased in patients with nasopharyngeal carcinorma compare with before radiotherapy (p<0.05, FDR correction, corrected, clusters=100mm3). No GM volume increased.Conclusions:The significantly increase of right hippocampus volume may be attributed to functional compensatory hypertrophy under the condition of temporal white matter injury, or damage of grey matter and inflammatory reaction. Radiotherapy may also have an effect on gray matter. High resolution imaging and VBM-DARTEL can detect radiation induced grey matter injury.更多还原