Scientific and Statistical Computing Core

National Institute Of Mental Health

Investigators

Linked publications & trials

Abstract

The principal mission of the Core is to help NIH researchers with analyses of their fMRI (brain activation mapping) and structural MRI (brain anatomy) data. Along the way, we also help non-NIH investigators, many in the USA but also some abroad. Several levels of help are provided, from short-term immediate aid to long-term development and planning. Consultations: The shortest-term help comprises in-person consultations with investigators about issues that arise in their research. These are quite varied, since there are many steps in carrying out fMRI and MRI data analyses and many different types of experiments. Common problems include: - How to set up experimental design so that data can be analyzed effectively. - Interpretation and correction of MRI imaging artifacts (for example: participant head motion during scanning; image warping due to magnetic field anomalies). - How to set up time series analysis to extract brain activation effects of interest and suppress non-activation imaging artifacts (e.g., from breathing). - How to analyze data to reveal connections between brain regions during specific mental tasks or at rest. - How to recognize poor quality data. - How to carry out reliable inter-patient (group) statistical analysis, especially when non-MRI data (e.g., genetic information, age, disease rating) needs to be incorporated. - How to get good alignment between the functional results and the anatomical reference images, and between the brain images from different participants. - What sequence of programs is "best" for analyzing a particular kind of data. - Reports of real or imagined bugs in the AFNI software, as well as feature requests (small, large, extravagant). - Analysis problems related to diffusion-weighted MRI data, which are acquired to reveal anatomical connections in the brain. There are familiar themes in many of these consultations, but each meeting and each experiment raises unique questions that require digging into the goals and details of the research project to ensure that nothing critical is overlooked. The first question asked by a user is often not the right question at all. When complex statistical or data-processing issues are raised, software often needs to be developed or modified to help researchers answer their specific questions. Helping with the Methods sections of papers, or with responses to reviewers, is also a part of our duties. Educational Efforts: The Core has developed (and updated) a 25-40-hour hands-on course on how to design and analyze fMRI data. We teach this course in various ways (in person, virtually, and as a hybrid of the two) at NIH and institutions around the US and world. All material for this continually evolving course (software, sample data, scripts, PDF slides, captioned videos) are freely available on our Web site (https://afni.nimh.nih.gov/pub/dist/doc/htmldoc/). The course material includes sample datasets, used to illustrate the entire process, starting with images output by MRI scanners and continuing through to the collective statistical analysis of data from groups of participants. More than 1000 AFNI forum postings were made by Core members, mostly in answer to queries from users. Algorithm and Software Development: The longest-term support consists of developing (or adapting) new methods and software for MRI data analysis, both to solve current problems and in anticipation of new needs. All of our software is incorporated into the AFNI package, which is Unix/Linux/Macintosh-based, open source, and available for download by anyone in source code (GitHub) or binary formats (Core server). New programs are created, and old programs modified, in response to specific user requests and in response to the Core's vision of what will be needed in the future. The Core also assists NIH labs in setting up computer systems for use with AFNI and maintains an active Web site with a forum for questions (and answers) about analysis of (f)MRI data, structural FMRI, and diffusion-based MRI. Notable developments during FY 2025 include: - Working with Drs. Wanyong Shin and Mark Lowe (Cleveland Clinic), we updated a tool called SLOMOCO for estimating and reducing motion effects in FMRI. Participant motion while being scanned often leads to tricky artifacts and is difficult to fully account for in processing. This tool makes use of a large number of AFNI programs, and estimates slicewise motion parameters. We showed how the updates lead to greater removal of residual motion artifacts in processed FMRI data. - We published a guide to FMRI processing with AFNI tools. This highlighted the wide range pipeline options that users can use to tailor their analysis designs to their study questions (which also vary quite widely across the field). It also showed the multiple layers of built-in provenance and quality control features, which should help users be as confident as possible about the processing steps they are using, the results of each step and the suitability of the data for their study. This manual also includes many general processing tips, for anyone setting up FMRI analysis. - We taught several educational FMRI training “Bootcamps”, including at NIH campus (USA), at the Laureate Institute for Brain Research (USA), and at the Indian Society of Neuroradiology hosted at NIMHANS (India). - We collaborated with Drs. Kadharbatcha Saleem and Peter Basser (NICHD) in developing both a subcortical atlas of the human brain (SAHB) and a Human Cerebellar Atlas (HCA) as reference datasets for the neuroscience community. These atlases represent some of the highest detailed delineations of the human brain with important applications to cerebellar diseases like ataxia and motor control disorders. - In collaboration with Drs. Yuan Zhong and Jian Kang (University of Michigan), we have extended our original region-based hierarchical modeling framework to the voxel level. This allows for improved modeling for analyzing a full brain dataset, compared to older massive univariate analysis approaches that use ad hoc multiple comparisons corrections. The project is nearing completion and will soon be integrated into a dedicated AFNI program for the neuroimaging community to use. - We are drafting a manuscript that highlights the benefits of effect quantification and robust meta-analytic methods for improving reproducibility in neuroimaging. Based on real data comparisons, the project has led to the development of a new AFNI tool stouffer. Public Health Impact: From Oct 2024 to Aug 2025, the principal AFNI publication has been cited in 727 works (cf Google Scholar). Most of our work supports basic research into brain function, but some of our work is more closely tied to or applicable to specific diseases: - We contributed FMRI analysis to a study on the relationship of inflammatory markers and resting state functional MRI (rsFMRI) in collaboration with Dr. Boadie Dunlop (Emory University School of Medicine) and Drs. Ki Sueng Choi and Helen Mayberg (both at the Icahn School of Medicine at Mount Sinai). This study examined the chronicity of major depressive disorder’s (MDD) relationship along with inflammation markers to rsFMRI as a predictor of long term outcomes. - We have continued work on a collaborative project with Dr. Lysianne Beynel (Noninvasive Neuromodulation Unit, NIMH) that combines realtime neurofeedback analysis with AFNI and repetitive transcranial magnetic stimulation (rTMS) of the amygdala (in collaboration with Dr. Vinai Roopchansingh, FMRI Core, NIMH). - We provided statistical assistance on analyzing the association of schizophrenia risk-associated SNPs and particular gene regions, in collaboration with Drs. Stefano Marenco (Human Brain Collection Core, NIMH) and Francis J McMahon (Human Genetics Brain, NIMH). The importance and potential pitfalls of covariate selection played a major role in the data analysis reported in the associated paper, which was published in Human Molecular Genetics. - Dilated perivascular spaces (PVSs) are associated with dementia and other related conditions. We collaborated with Dr. Alan Koretsky (Laboratory of Functional and Molecular Imaging, NINDS) and colleagues on a high-field project to map PVSs at high resolution, and to investigate PVS morphology. This helps lay groundwork for using PVSs in diagnostic or prognostic applications in neurological diseases. - We helped on a study for quantifying structural changes in pediatric patients who had undergone surgery for cortical resection or ablation, led by Dr. Marlene Berhmann and colleagues (University of Pittsburgh). This study found, in part, that there is a hemispheric difference in structural plasticity, which may have implications for clinical outcomes.