Spurious model comparisons are widespread in biomedical artificial intelligence
Tianchu Zeng, Hetu Li, Shaoshi Zhang, Yan Quan Tan, Fang Tian, Csaba Orbán, Lijun An, Wanyu Che, Jingwen Cheng, Joanna Su Xian Chong, Niousha Dehestani, Zijian Dong, Xin Li, Zhizhou Li, Mervyn Jun Rui Lim, , B.T. Thomas Yeo
97% of 184 high-impact biomedical AI studies compare cross-validated models with tests assuming fold independence; in 69% of re-analyzable studies a significant comparison does not survive correction.
bioRxiv · Sep 10, 2026 · PreprintBioinformatics✦ In-depth digest