5. Sites, centers and collaborators¶
Partner centers¶
The lab's data and much of its day-to-day operation come from clinical partners. For each, the wiki should eventually answer the same five questions: who our contacts are, what data flows in which direction, which systems hold it, under which agreement, and which procedures apply. Today only the CNT column is complete.
| CNT / Penn Epilepsy Center | Penn Stroke Center · LCNS · brainSTIM | Center for Brain Injury and Repair (CBIR) | IBI · DBEI · CCEB | |
|---|---|---|---|---|
| Relationship | Shared servers, coordinators and patients; lab touchdown space | Collaboration on stroke, aphasia and neuromodulation cohorts | Collaboration on TBI cohorts and imaging | Academic home; compute, training, methods collaborators |
| Data in | iEEG, scalp EEG, MRI/CT, RNS, surgical outcomes, REDCap | to document: lesion imaging, behavioural/outcome scales, stimulation protocols | to document: TBI imaging, clinical trajectories | none (methods, not data) |
| Data out | Derivatives, de-identified datasets, Pennsieve/ieeg.org uploads | to document | to document | manuscripts, shared code |
| Systems | cnt1, cnt-fs, BSC, Natus, Flywheel, REDCap, ieeg.org, Pennsieve | to document | to document | PMACS, LPC |
| Agreements | Penn IRB protocols (see Operations › Regulatory) | to document: IRB / reliance / DUA | to document | n/a |
| Contacts | PIs: Erin Conrad, Kate Davis, Brian Litt; lead CRC | to add | to add | to add |
| Procedures in this wiki | Electrophysiology, Imaging, REDCap, Operations | none yet | none yet | Compute › PMACS, Operations › Onboarding |
When a new center starts sending data, create its procedures in the existing themes (an Imaging pipeline is an Imaging pipeline whichever center the scan came from) and tag the pages with program: stroke, program: tbi, and so on, so the Map can later show a per-program view.
The site network¶
Beyond the Penn centers, the lab curates data from 15+ external sites. The inventory — what each site provides (iEEG, imaging, RNS, metadata), counts, whether a protocol or DUA is required, and where the data lands — is maintained with the Data theme. (Action: turn that list into a table with columns site · modalities · n · protocol/DUA · location · status · contact, and keep it in this wiki.)
Onboarding a new site — the protocol¶
This is the two-year goal from the data-coordinator role made into a checklist. Every new site, and every new partner center, goes through it in order.
- Agreement — confirm whether the data needs a protocol, reliance agreement or DUA; record the decision and the document on the site's page.
- Transfer channel — agree the route (Pennsieve, Box, SFTP, drive) and who sends. Identified data may land only on approved PMACS systems (cnt1, cnt-fs, BSC). Use the agreed project destination and access controls.
- Intake — log the delivery: date, contents, counts, sender. Check against what was promised.
- Conformance — restructure to the lab's data structure specification (Data Structure v1.0, being written into Data › Standards); record what could not be mapped.
- De-identification — run the Data › De-identification scripts; verify headers.
- Registration — add or update the site in the inventory; note derivatives available.
- Feedback to the site — what was missing (the current inventory already flags gaps such as missing outcomes or iEEG reconstruction).
Contacts by site¶
The site table is in preparation.