PrimateX: a unified multi-agentic framework for standardized MRI analysis across primate species

Abstract Standardized preprocessing has transformed human neuroimaging into a reproducible, population-scale enterprise, yet equivalent computational frameworks for non-human primates (NHPs) remain absent, forcing continued reliance on fragmented, labor-intensive workflows. Here we introduce PrimateX, a multi-agentic computational framework in which LLM-guided orchestration dynamically routes structural MRI scans through species-tailored, topology-preserving workflows, extracting cortical surfaces directly from raw volumes with full provenance tracking across marmosets, macaques, chimpanzees, and humans. Validated on 1,221 NHP and 1,113 human scans, PrimateX outperformed existing NHP pipelines in segmentation and surface reconstruction while matching dedicated human tools such as FSL in their native domain. Through MRI-style data augmentation, the framework achieved robust zero-shot generalization to 8 unseen primate species across 31 cohorts from 19 global sites (n = 157). Critically, the routing architecture and closed-loop quality-control mechanisms, demonstrated through diverse downstream neuroimaging analyses and cross-species research, establish PrimateX as a widely-applicable computational paradigm. PrimateX is available at https://primatex.pku.edu.cn.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-17
DOI
https://doi.org/10.21203/rs.3.rs-11234334/v1
Primary Topic
Medical Image Segmentation Techniques
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

PrimateX: a unified multi-agentic framework for standardized MRI analysis across primate species

Huazhu Fu, 蔡程韬, Sze Chai Kwok, Yunfeng Wang et al.
1 links
Research Square
Medical Image Segmentation Techniques
preprint

PrimateX: a unified multi-agentic framework for standardized MRI analysis across primate species

Huazhu Fu, 蔡程韬, Sze Chai Kwok, Yunfeng Wang, Ching‐Yu Cheng, Alejandro Federico Frangi, Ronghui Ju, Stefan Everling, Hong Liang, Junichi Hata, 崔凯龙, Daisuke Yoshimaru, Ken Nakae, Haoran Zhai, Zheng Wang, Shan Cong, Shiqi Yu, Yiheng Hu, Chenyuan Bian, Ravi Menon, Yichun Huang, Hongyi Li, Xiaohui Yao, David Leopold, Jing Liu, Jiahao Huang, Hideyuki Okano, Chet Sherwood, Yuanjing Feng, Junlong Huang, William Hopkins
preprint en

Abstract

Abstract Standardized preprocessing has transformed human neuroimaging into a reproducible, population-scale enterprise, yet equivalent computational frameworks for non-human primates (NHPs) remain absent, forcing continued reliance on fragmented, labor-intensive workflows. Here we introduce PrimateX, a multi-agentic computational framework in which LLM-guided orchestration dynamically routes structural MRI scans through species-tailored, topology-preserving workflows, extracting cortical surfaces directly from raw volumes with full provenance tracking across marmosets, macaques, chimpanzees, and humans. Validated on 1,221 NHP and 1,113 human scans, PrimateX outperformed existing NHP pipelines in segmentation and surface reconstruction while matching dedicated human tools such as FSL in their native domain. Through MRI-style data augmentation, the framework achieved robust zero-shot generalization to 8 unseen primate species across 31 cohorts from 19 global sites (n = 157). Critically, the routing architecture and closed-loop quality-control mechanisms, demonstrated through diverse downstream neuroimaging analyses and cross-species research, establish PrimateX as a widely-applicable computational paradigm. PrimateX is available at https://primatex.pku.edu.cn.

Research Square
University of Fukui (JP), Agency for Science, Technology and Research (SG), Western University (CA), National Institutes of Health (US), The University of Texas MD Anderson Cancer Center (US), Harbin Engineering University (CN), National University of Singapore (SG), George Washington University (US), Peking University (CN), Keio University (JP), Hainan University (CN), RIKEN Center for Brain Science (JP), University of Manchester (GB), Liaoning Provincial People's Hospital (CN), Affiliated Hospital of Qingdao University (CN), Peking University First Hospital (CN), Duke Kunshan University (CN), Zhejiang University of Technology (CN)
Medical Image Segmentation Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon