摘要
目的纹理分析是一种新型的图像组学技术,通过在图像上选择特定的感兴趣区域,提取和分析该区域的纹理特征参数,可以辅助判断疾病(尤其是肿瘤)的特征和预后。肺癌是我国发病率较高的肿瘤之一,目前肺癌的纹理分析研究大多选择肿瘤原发灶作为感兴趣区域,通过MaZda、Matlab或自研软件获取纹理特征参数,从而判断肿瘤病理类型、基因表达、治疗反应以及预后等。由于相关研究较少,且纹理特征提取及分析方法不一致,尚未得出统一的结论。但纹理分析作为一种成本低、无创伤的方法,在判断肿瘤临床特征及预后方面仍具有较高的临床应用价值,可为今后指导临床实践提供重要依据。在现有研究的基础上,未来的研究可以尝试选择多种感兴趣区域,比较不同的特征提取和分析方法,同时增大样本含量,从而提高研究的可信度。
Texture analysis is a new type of image omics technology.By extracting and analyzing the texture feature parameters of the specific region of interest on an image,it can assist in judging the characteristics and prognosis of diseases(especially tumors).Lung cancer is the most common tumor in China.At present,most of the texture analysis studies on lung cancer selected the primary tumor as the region of interest,and obtained texture feature parameters through MaZda,Matlab or self-developed software,so as to determine the tumor pathological type,gene expression,treatment response and prognosis.Due to few related studies and inconsistent texture feature extraction and analysis methods,a unified conclusion has not yet been drawn.However,as a low-cost and non-invasive method,texture analysis still has high clinical application value in judging the clinical characteristics and prognosis of tumors,and can provide an important basis for guiding clinical practice in the future.On the basis of the current research,future research can try to select a variety of regions of interest,compare different feature extraction and analysis methods,and increase the sample size,thereby improving the credibility of the research.
作者
王文婷
李光
WANG Wen-ting;LI Guang(Department of Radiotherapy,First Hospital of China Medical University,Shenyang 110000,China)
出处
《中华肿瘤防治杂志》
CAS
北大核心
2022年第4期235-239,共5页
Chinese Journal of Cancer Prevention and Treatment
关键词
纹理分析
肺癌
病理类型
临床分期
预后
texture analysis
lung cancer
pathological type
clinical stages
prognosis
