UNASSIGNED: The authors retrospectively retrieved 4542 text reports and images for fluorodeoxyglucose PET/CT lymphoma examinations from 2008 to 2018 in the University of Wisconsin-Madison institutional clinical imaging database. Of these total reports, 1664 had Deauville scores that were extracted from the reports and served as training labels. The bidirectional encoder representations from transformers (BERT) model and initialized BERT models BioClinicalBERT, RadBERT, and RoBERTa were adapted to the nuclear medicine domain by pretraining using masked language modeling. These domain-adapted models were then compared with the non-domain-adapted versions on the task of five-point Deauville score prediction. The language models were compared against vision models, multimodal vision-language models, and a nuclear medicine physician, with sevenfold Monte Carlo cross-validation. Means and SDs for accuracy are reported, with P values from paired t testing.
UNASSIGNED: Domain adaptation improved the performance of all language models (P = .01). For example, BERT improved from 61.3% ± 2.9 (SD) five-class accuracy to 65.7% ± 2.2 (P = .01) following domain adaptation. Domain-adapted RoBERTa (named DA RoBERTa) performed best, achieving 77.4% ± 3.4 five-class accuracy; this model performed similarly to its multimodal counterpart (named Multimodal DA RoBERTa) (77.2% ± 3.2) and outperformed the best vision-only model (48.1% ± 3.5, P ≤ .001). A physician given the task on a subset of the data had a five-class accuracy of 66%.
UNASSIGNED: Domain adaptation improved the performance of large language models in predicting Deauville scores in PET/CT reports.Keywords Lymphoma, PET, PET/CT, Transfer Learning, Unsupervised Learning, Convolutional Neural Network (CNN), Nuclear Medicine, Deauville, Natural Language Processing, Multimodal Learning, Artificial Intelligence, Machine Learning, Language Modeling Supplemental material is available for this article. © RSNA, 2023See also the commentary by Abajian in this issue.
■作者回顾性检索了2008年至2018年在威斯康星大学麦迪逊分校机构临床影像学数据库中进行的4542份氟脱氧葡萄糖PET/CT淋巴瘤检查的文本报告和图像。在这些总报告中,1664年,从报告中提取了多维尔分数,并用作培训标签。来自变压器(BERT)模型和初始化BERT模型的双向编码器表示BioClinicalBERT,拉德伯特,和RoBERTa通过使用蒙面语言建模进行预训练而适应了核医学领域。然后在五点Deauville分数预测的任务中,将这些域适应的模型与非域适应的版本进行比较。将语言模型与视觉模型进行了比较,多模态视觉语言模型,和核医学医生,七倍蒙特卡罗交叉验证。报告了用于准确性的均值和SDs,用配对t检验的P值。
■域自适应提高了所有语言模型的性能(P=.01)。例如,领域适应后,BERT从61.3%±2.9(SD)五级精度提高到65.7%±2.2(P=0.01)。领域改编的RoBERTa(名为DARoBERTa)表现最好,达到77.4%±3.4五级精度;该模型的性能与多模式对应物(称为多模式DARoBERTa)相似(77.2%±3.2),优于最佳仅视觉模型(48.1%±3.5,P≤.001)。对数据子集进行任务的医生具有66%的五级准确率。
■领域适应提高了大型语言模型在预测PET/CT报告中的Deauville分数方面的性能。关键词淋巴瘤,PET,PET/CT,迁移学习,无监督学习,卷积神经网络(CNN)核医学,多维尔,自然语言处理,多模态学习,人工智能,机器学习,语言建模补充材料可用于本文。©RSNA,202另见本期Abajian的评注。