Cross-site prognosis prediction for nasopharyngeal carcinoma from incomplete multi-modal data

Chuan-Xian Ren, Gengxin Xu, Dao-Qing Dai, Li Lin, Ying Sun, Qing-Shan Liu

Abstract: Highlights•We study the cross-site prognosis prediction problem for nasopharyngeal carcinoma.•We propose a cross-attention multi-modal fusion module to learn site-adaptive space.•We design cross-domain partial contrastive learning to improve discriminant ability.•Our method obtains superior performance in cross-site prognosis prediction experiment.•This research has a significant value in clinical practice for its site-adaptability.