【报告摘要】 The rfGWAS method addresses challenges in brain imaging genetic studies by focusing on small regions, mitigating misalignment and registration errors. Its streamlined computation preserves meaningful associations, confirmed by theoretical analysis adhering to the independence-zero equivalence principle. Simulations demonstrate effectiveness, and application to ADNI study hippocampus surface data showcases potential.

【报告人简介】中国科学院数学与系统科学研究院副研究员、博士生导师。主要从事统计学习、度量空间非参数方法等方面研究，在AoS、JASA等统计学杂志发表论文20余篇，荣获2022年教育部高等学校科研优秀成果自然科学类二等奖（排名第二）。