EVENT DOSSIER
On September 7, 2026, researchers released the Real-World Multi-Modal and Longitudinal Lung Cancer Dataset. This dataset includes whole-slice images, CT scans, and PET scan images of 1,365 lung cancer patients, as well as structured clinical data, transcriptomic data, and longitudinal treatment information. There are significant cross-modal missing values in the data. The study provided single-modal and multi-modal benchmark tests for predicting 12-month overall survival, disease-specific survival, and risk under severe missing data conditions. Results showed that integrating complementary modalities improved prediction performance more than using a single-modal approach. The dataset and benchmark code have been released on the GitHub platform.
研究人员发布了 Real-World Multi-Modal and Longitudinal Lung Cancer Dataset，该数据集包含 1,365 名肺癌患者数据。数据集涵盖三种影像模态（全切片图像、CT 扫描和 PET 扫…
研究人员发布了 Real-World Multi-Modal and Longitudinal Lung Cancer Dataset，该数据集包含 1,365 名肺癌患者数据。数据集涵盖全切片图像、CT 扫描和 PET 扫描三种影像模态…
研究人员发布了 Real-World Multi-Modal and Longitudinal Lung Cancer Dataset，该数据集包含 1,365 名肺癌患者数据。数据集涵盖三种影像模态（全切片图像、CT 扫描和 PET 扫描）、结构化临床数据、转录组学数据以及纵向随访和治疗信息，并存在显著的跨模态非均匀缺失。研究提供了针对 12 个月总生存期预测、疾病特异性生存及严重缺失数据下风险预测的单一模态和多模态基准测试。结果显示，尽管存在高比例缺失，整合互补模态比单一模态方法能一致地提升预测性能。数据集和基准代码已发布在 https://github.com/ritacmendes/MMIST-LUNG。
研究人员发布了 Real-World Multi-Modal and Longitudinal Lung Cancer Dataset，该数据集包含 1,365 名肺癌患者数据。数据集涵盖全切片图像、CT 扫描和 PET 扫描三种影像模态，以及结构化临床数据、转录组数据和纵向随访治疗信息，并存在显著的跨模态非均匀缺失。研究提供了针对 12 个月总生存期预测、疾病特异性生存及严重缺失数据下风险预测的单一模态与多模态基准测试。结果显示，尽管存在高比例缺失，整合互补模态比单一模态方法更能提升预测性能。数据集和基准代码已发布在 https://github.com/ritacmendes/MMIST-LUNG。