社交媒体情感分析揭示公众对CT结肠成像与结肠镜的认知差异
本研究通过情感分析比较公众对CT结肠成像(CTC)与光学结肠镜的认知。收集2015年1月至2021年9月的英文推文,使用Twitter-roBERTa-base模型进行情感评分。结果显示CTC相关推文量仅为结肠镜的约1%,但CTC负面推文比例(9.81%)低于结肠镜(31.8%),提示公众对CTC负面感知较少,可能有助于推广其作为结直肠癌筛查替代方法。
要点
- CTC相关推文量仅为结肠镜的约1%,公众认知有限。
- CTC负面推文比例(9.81%)明显低于结肠镜(31.8%)。
- CTC正面推文比例(21.63%)高于结肠镜(16.9%)。
局限性
- 仅分析社交媒体推文,不能代表一般人群观点。
- 推文情感分析模型可能存在分类偏差。
- 未报告CTC与结肠镜推文量的具体统计比较方法。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
PURPOSE: The purpose of this study was to understand the public perception of CT colonography (CTC) in comparison with optical colonoscopy as a colorectal cancer screening technique.
METHODS: In this observational study, all English-language tweets from January 1, 2015, until September 1, 2021, containing terms related to CTC and terms related to optical colonoscopy were collected. The tweets were given sentiment scores using Twitter-roBERTa-base, a natural language processing model. These scores were then used to classify tweets into positive, neutral, and negative categories. The numbers of negative, positive, and neutral tweets were tabulated.
RESULTS: A total of 4,709 tweets from 2,194 users relating to CTC were collected. Of these tweets, 9.81% were negative, 68.52% were neutral, and 21.63% were positive. In comparison, a total of 445,969 tweets from 261,209 users were collected relating to optical colonoscopy. Of these tweets, 31.8% were negative, 51.3% were neutral, and 16.9% were positive.
CONCLUSIONS: The public awareness of CTC remains limited in comparison with optical colonoscopy, with Twitter volume relating to CTC being about 1% the volume for optical colonoscopy. There was a higher proportion of negative tweets regarding colonoscopy. The lower proportion of negative tweets regarding CTC may be helpful in encouraging its use as an alternative to optical colonoscopy, with the aim of increasing uptake of colorectal cancer screening.
引用
- Chen, LeBedis, Chang, et al. The Public Perception of CT Colonography Versus Colonoscopy via Sentiment Analysis of Social Media. J Am Coll Radiol. 2023. PMID: 37127218. doi:10.1016/j.jacr.2023.03.011
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。