Brain MRI research often involves more than reviewing one study at a time. Research teams may need to organize hundreds of cases, account for multiple imaging sequences, preserve case-specific context, monitor processing, and review results consistently across an entire cohort.
To support these workflows, we are introducing Large-Scale Analysis in ReMIND: a batch-processing workspace designed for approved academic and institutional research projects.
From individual studies to organized cohort review
ReMIND's standard workflow allows users to upload and review an individual brain MRI study. Large-Scale Analysis extends that experience to multi-case datasets while keeping every study organized as a separate case.
Approved research teams can:
- Create a dedicated analysis job for a project.
- Add individual cases manually or import multiple cases from a structured folder or ZIP archive.
- Include multiple MRI sequences or volumes for each case.
- Provide optional case-specific clinical context.
- Submit the job for background processing.
- Monitor progress and return to an active job later.
- Review each completed case with its MRI sequences, generated report draft, and case-specific follow-up chat.
The goal is not simply to generate reports in bulk. It is to provide researchers with a consistent workspace for moving from dataset preparation to case-level review without losing the structure of the original cohort.
Designed around common research datasets
Large-Scale Analysis supports datasets in which each case is stored in its own folder. A case may contain one or more NIfTI volumes or DICOM series, including sequences such as T1, T2, FLAIR, DWI, SWI, post-contrast imaging, and angiographic studies.
Clinical context is optional. When relevant to the research question, it can be entered manually, included as a text file within the case folder, or supplied through a CSV or XLSX mapping file.
This structure allows teams to preserve their own Case IDs and maintain a clear relationship between the source study, any available context, and the resulting ReMIND output.
Your data stays your data
Uploaded studies are used only to complete the analysis job requested by the approved research team.
Files and generated outputs are retained only according to the configured job-retention period and are stored within protected application storage while the job remains active. Research teams remain responsible for appropriate de-identification, permitted data use, institutional approval, and compliance with their own data-governance requirements.
Because metadata processing cannot guarantee the removal of identifiers embedded directly within image pixels, submitted studies should be appropriately reviewed and de-identified before upload.
Who is Large-Scale Analysis for?
The workflow is intended for:
- Academic research laboratories studying brain MRI cohorts.
- Imaging and clinical research groups conducting retrospective studies.
- Teams evaluating reporting workflows or vision-language models.
- Institutional projects with appropriate oversight and data-governance procedures.
Access is reviewed based on project fit, available processing capacity, and data-handling requirements. Applying does not automatically guarantee access.
A research tool with clear boundaries
ReMIND remains a research prototype and is not approved for clinical diagnosis.
Generated reports are drafts intended for review by qualified clinicians or researchers. Performance may vary across scanners, imaging protocols, institutions, and patient populations. ReMIND is designed to support structured research review, not replace clinical interpretation.
Interested in using ReMIND for a research project?
Researchers can learn more about the supported workflow, dataset preparation, privacy practices, and access requirements on the Large-Scale Analysis page.