东篱科研大数据发现系统（DRDS）

位置：成果数据库 > 期刊 > 期刊详情页

基于PET图像纹理分析食管癌~(18)F-FDG摄取异质性

ISSN号：1003-3289
期刊名称：中国医学影像技术
时间：2013.10.6
页码：1036-1040
分类：R735.1[医药卫生—肿瘤;医药卫生—临床医学] R817.4[医药卫生—影像医学与核医学;医药卫生—放射医学;医药卫生—临床医学]
作者机构：[1]山东省肿瘤医院放疗科,山东济南250117, [2]济南大学山东省医学科学院生命与科学学院,山东济南250022, [3]山东省肿瘤医院PET/CT中心,山东济南250117, [4]山东师范大学物理与电子科学学院,山东济南250004
相关基金：国家自然科学基金（81101700）
相关项目：多尺度分析技术在图像引导肿瘤精确放疗中的研究及临床应用

关键词：食管肿瘤, 正电子发射型体层摄影术, 18F氟脱氧葡萄糖, Esophageal neoplasms, Positron-emission tomography, Fluorodeoxyglucose F18

中文摘要：

目的提取食管癌原发灶PET图像纹理特征,并量化肿瘤18 F-FDG摄取异质性,探讨食管癌18 F-FDG摄取异质性与最大标准摄取值（SUVmax）及病理参数的关系。方法回顾性分析30例术前接受18 F-FDG PET/CT全身扫描的食管癌患者。应用Matlab 7.6软件计算食管癌原发灶PET图像纹理参数（对比度、相关性、熵、能量）,分析各纹理参数与SUVmax、肿瘤浸润深度、分化程度及淋巴结转移情况的相关性。结果食管癌原发灶纹理参数均与SUVmax相关,分别与对比度、熵呈正相关（r=0.537,P=0.002;r=0.434,P=0.017）,与相关性、能量呈负相关（r=-0.471,P=0.009;r=-0.450,P=0.012）;在不同浸润深度和淋巴结转移情况下,纹理参数熵和能量的组间差异有统计学意义（P均〈0.05）,并分别与熵呈正相关（rs=0.574,P=0.001;rs=0.366,P=0.047）,与能量呈负相关（rs=-0.428,P=0.018;rs=-0.436,P=0.016）。各代谢参数与食管癌分化程度无相关性。结论纹理参数可量化18F-FDG摄取异质性,提供能反映肿瘤生物学特征的丰富影像学信息。

英文摘要：

Objective To quantify ^（18）F-FDG uptake heterogeneity of esophageal carcinoma using PET image texture features, and to explore the correlation with maximum standardized uptake value （SUVmax） and pathological parameters. Methods Thirty patients with esophageal carcinoma underwent whole body ^（18）F-FDG PET/CT scan before surgical operation were analyzed retrospectively. To quantify TM F-FDG uptake heterogeneity on pre-treatment 18 F-FDG PET images, four texture features （contrast, correlation, entropy and energy） were extracted using Matlab 7.6 software algorithm. The texure parameters were correlated with SUVmax, depth of invasion, differentiation of the primary lesions and lymph metastasis status. Results The primary esophageal tumors with high SUVmax were more heterogeneous on uptake. Correlation was found in contrast （r=0. 537, P=0. 002）, correlation （r=-0. 471, P=0. 009）, entropy （r=0. 434, P=0. 017） and energy （r=- 0. 450, P = 0. 012）, respectively. Depth of invasion was correlated with entropy （r3 = 0.574, P = 0.001 ） and energy （r3 =- 0. 428, P= 0.018）. There were also statistically significant differences in entropy （r3 = 0.366, P = 0. 047） and energy （r3 =- 0. 436, P= 0. 016） between the groups with or without lymph node metastasis. There was no significant correlation between texture features and degree of differentiation. Conclusion Tumor is F-FDG uptake heterogeneity quantified by texture features has potential to provide more functional image information on biological characteristics.

同期刊论文项目