Division of Medical Information Sciences, Geneva University Hospital, Switzerland.
Department of Radiology and Medical Informatics, University of Geneva, Switzerland.
Stud Health Technol Inform. 2022 May 25;294:849-853. doi: 10.3233/SHTI220605.
The present study shows first attempts to automatically classify oncology treatment responses on the basis of the textual conclusion sections of radiology reports according to the RECIST classification. After a robust and extended manual annotation of 543 conclusion sections (5-to-50-word long), and after the training of several machine learning techniques (from traditional machine learning to deep learning), the best results show an accuracy score of 0.90 for a two-class classification (non-progressive vs. progressive disease) and of 0.82 for a four-class classification (complete response, partial response, stable disease, progressive disease) both with Logistic Regression approach. Some innovative solutions are further suggested to improve these scores in the future.
本研究首次尝试根据 RECIST 分类，基于放射学报告的文本结论部分自动分类肿瘤治疗反应。在对 543 个结论部分（5 到 50 个单词长）进行了稳健且广泛的人工标注后，并对几种机器学习技术（从传统机器学习到深度学习）进行了训练后，最佳结果显示 Logistic Regression 方法的二类分类（非进展性与进展性疾病）的准确率为 0.90，四类分类（完全缓解、部分缓解、稳定疾病、进展性疾病）的准确率为 0.82。进一步提出了一些创新的解决方案，以提高未来的这些分数。
