Related Experiment Video
Updated: Sep 27, 2025
A Standardized Pipeline for Examining Human Cerebellar Grey Matter Morphometry using Structural Magnetic Resonance Imaging
Published on: February 4, 2022
A model-based approach to assess reproducibility for large-scale high-throughput MRI-based studies.
Zeyu Jiao1, Yinglei Lai2, Jujiao Kang1
1Shanghai Center for Mathematical Sciences, Fudan University, 220 Handan Road, Shanghai, China; Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China; Key Laboratory of Computational Neuroscience and Brain-Inspired Intelligence (Fudan University), Ministry of Education, China; MOE Frontiers Center for Brain Science, Fudan University, Shanghai, China; Zhangjiang Fudan International Innovation Center, China.
A new model-based index quantifies the reproducibility of neuroimaging studies. This method confirms high reproducibility in large studies and helps determine optimal sample sizes for reliable Magnetic Resonance Imaging (MRI) findings.
Area of Science:
- Neuroscience
- Medical Imaging
- Biostatistics
Background:
- Magnetic Resonance Imaging (MRI) is vital in neuroscience research.
- Reproducibility of MRI findings, especially in association and activation studies, is under debate.
- Current reproducibility measures are limited and do not assess overall study reliability.
Purpose of the Study:
- To propose a model-based reproducibility index for large-scale MRI studies.
- To quantify reproducibility in association and task-induced brain activation studies.
- To establish a relationship between sample size and study reproducibility.
Main Methods:
- Developed a model-based reproducibility index for high-throughput MRI studies.
- Assessed reproducibility using large sMRI/fMRI databases (e.g., UK Biobank, HCP).
- Created a tool to determine minimal sample size for desired reproducibility.
Main Results:
- Demonstrated high model-based reproducibility (>0.99) for large sample size association studies (e.g., brain structure/function vs. BMI).
- Observed similar high reproducibility for Motor Image Dataset (MID) task activation.
- Identified that both sample size and study-specific factors influence reproducibility.
Conclusions:
- A systematic assessment of reproducibility is crucial for current large-scale MRI studies.
- The proposed index provides a reliable measure for overall study reproducibility.
- The analytical tool aids in experimental design by defining necessary sample sizes.
