基于文献计量学的近5年放射组学研究热点分析

Bibliometric and research hotspots analysis of radiomics in recent five years

目的：从文献计量学角度，探讨近5年放射组学的研究热点。方法下载Pubmed数据库中近5年发表的放射组学研究文献题录，使用Bicomb 2.0软件统计分析文献的发表年代、来源期刊及期刊所属国家、第一作者以及高频主要主题词分布情况。对文献第一作者所属研究机构进行人工逐条统计分析。将出现频次不小于2次的主要主题词作为高频主要主题词，建立词篇矩阵，使用SPSS 22.0软件进行聚类分析，得到该领域研究热点。结果共检索到文献56篇，其中2012、2013、2014、2015、2016年分别为2（3.6%，2/56）、3（5.4%，3/56）、4（7.1%，4/56）、28（50.0%，28/56）和19篇（33.9%，19/56）。56篇文献来源于10个国家的34种期刊，将载文量不小于3篇的期刊划分为核心区，入选期刊共5种，共21篇文献（37.5%，21/56）。34种期刊（56篇文献）频次分布分别为美国18种、英国6种、荷兰2种、德国2种，瑞士、法国、西班牙、爱尔兰、土耳其、印度各1种。56篇文献中，美国27篇、荷兰10篇、加拿大5篇，英国、中国、法国、捷克斯洛伐克各2篇，德国、西班牙、匈牙利、韩国、日本、印度各1篇。高频主要主题词共11个，凝聚为A、B两大类研究热点方向，其中热点A是结合三维后处理、计算机辅助等方法诊断非小细胞肺癌；热点B主要是PET中结合计算机辅助和影像整合等技术诊断肺肿瘤。结论近5年放射组学研究热点主要集中在结合三维后处理、计算机辅助等方法诊断非小细胞肺癌以及PET中结合计算机辅助和影像整合等技术诊断肺肿瘤2个领域。

Objective To investigate hotspots from researches on radiomics in recent five years. Methods Bibliographies from research literature of radiomics in recent five years in PubMed database were downloaded. The publication years, journals, countries of publication, the first authors and the frequency of Mesh major topics were counted by Bicomb 2.0 software. The affiliations were analyzed artificially. Mesh major topics appeared no less than two times were intercepted as high frequency ones and the high frequency Mesh major topics matrix were formed. SPSS 22.0 statistical software were applied for clustering analysis with matrix to get the topic hotspots. Results A total of 56 literatures were screened out, among which, there were 2 （3.6%,2/56）, 3 （5.4%,3/56） ,4 （7.1%,4/56） ,28 （50.0%,28/56） and 19 （33.9%, 19/56） literatures published in 2012, 2013, 2014, 2015, 2016, respectively. Twenty one of 56 （37.5%, 21/56） literatures published in 5 journals which were contained in 34 journals from 10 countries, were divided into core area, in terms of published articles of no less than 3. Frequency distribution were 18 （US）, 6 （England）, 2 （Netherland and Germany） and 1 （Switzerland, France, Spain, Ireland, Turkey, India）, respectively, for 34 journals （56 literatures）.Among the 56 literatures, US, Netherlands and Canada＆amp;nbsp;contained 27, 10, 5, respectively. England had 2, as same as China, France and Czech Republic, and 1 in Germany, Spain, Hungary, Korea, Japan, India, respectively, in terms of first authors affiliated research institution. The number of high frequency major topics was 11 and among them 2 research hotspots （A and B） were clustered. One （A） was computer-aided diagnosis of non-small cell lung cancer, combined with methods of three-dimensional processing, another （B） was PET combined with computer-aided and image integration technology in the diagnosis of lung tumors. Conclusion Hotspots from researches on radiomics in recent five years are focused on computer-aided diagnosis of non-small cell lung cancer, combined with methods of three-dimensional processing, and PET combined with computer-aided and image integration technology in the diagnosis of lung tumors.