发表机构
Ataraxis AI; Dana-Farber Cancer Institute; Harvard Medical School; Rutgers Cancer Institute; Northwell Health; University of Pittsburgh; Magee-Womens Hospital of UPMC; Medical University of Łódź; Shaare Zedek Medical Center; The Hebrew University of Jerusalem; University of Chicago; Jagiellonian University Medical College; Fred Hutchinson Cancer Center; University of Washington; NYU Grossman School of Medicine; Intelligible, Inc.(Ataraxis AI; 丹娜-法伯癌症研究所; 哈佛医学院; 罗格斯癌症研究所; 诺斯韦尔健康; 匹兹堡大学; UPMC麦奇妇女医院; 罗兹医科大学; 沙雷泽德克医疗中心; 耶路撒冷希伯来大学; 芝加哥大学; 雅盖隆大学医学院; 弗雷德·哈钦森癌症中心; 华盛顿大学; 纽约大学格罗斯曼医学院; Intelligible公司)
机构由 AI 辅助整理,请以论文原文为准。AI 中文总结
该研究提出两阶段AI模型,利用转录组信息从病理图像预测乳腺癌新辅助治疗反应,在1,412名患者中达到AUROC 0.79,优于传统病理标志物,展示了数据稀疏场景下的应用潜力。
AI 中文摘要
标记数据的稀缺限制了肿瘤学中深度学习生物标志物的发展。我们开发了一个两阶段人工智能模型,用于预测乳腺癌对新辅助治疗的病理完全缓解(pCR)。第一阶段利用来自32种癌症类型的8,742名患者,从组织病理学中学习转录组,并通过病理学家审查和与测量表达的空间一致性得到证实。这简化了第二阶段,即从推断的表达和临床变量预测pCR。该模型使用1,080名患者(五个队列)开发,并在1,412名患者(九个队列)中评估,实现了汇总AUROC为0.79(95%置信区间,0.73-0.85),能够在分子亚型内区分应答者。它优于组织病理学生物标志物,在肿瘤内采样和最小活检组织下保持稳定。消融研究表明,转录组范围的推断比单独使用临床变量或单阶段病理模型更能提高区分能力,并通过避免基因组检测的基因选择约束增强了稳健性。这些结果表明,生物学信息压缩可能推广到精准肿瘤学中数据稀疏的应用。
英文摘要
Scarcity of labeled data limits development of deep learning biomarkers in oncology. We develop a two-stage AI model predicting pathological complete response (pCR) to neoadjuvant therapy in breast cancer. The first stage learns the transcriptome from histopathology using 8,742 patients across 32 cancer types, corroborated by pathologist review and spatial agreement with measured expression. This simplifies the second stage to predicting pCR from inferred expression and clinical variables. Developed using 1,080 patients (five cohorts) and evaluated in 1,412 patients (nine cohorts), the model achieves a pooled AUROC of 0.79 (95% CI, 0.73-0.85), discriminating responders within molecular subtypes. It outperforms histopathological biomarkers, remaining stable across intratumoral sampling and with minimal biopsy tissue. Ablations show transcriptome-wide inference improves discrimination over clinical variables alone or one-stage pathology models, and robustness by avoiding genomic assays' gene selection constraints. These results indicate that biologically informed compression may generalize to data-sparse applications in precision oncology.