# Technical Monitoring (minutes) - 2023 Feb 14

:::info
**Date:** Feb 14, 2023 17:00(UTC)

**Agenda**: [link](https://hackmd.io/NvT4OCT2Tmi3c_SJTMQlig)

**Participants:**
- Elodie (ES)
- Chris (CM)
- Celine (CP)
- Mathias (MG)
- Eilidh (EM)
- Oscar (OE)

**Minutes by:** ES
:::

### Organization of the meetings (ES)
* Minutes and agendas of the meetings are now on Hackmd.
* Everybody can add topic suggestions for the agenda of the next meetings.
### Nibabies (MG)
* Shifting from FreeSurfer to [MIRTK](https://github.com/BioMedIA/MIRTK) for surface recon :
Initially aiming for [MCRIBS](https://github.com/DevelopmentalImagingMCRI/MCRIBS) but [MCRIBS] requires lots of specific preprocessing steps -> going back to [MIRTK].
* Topology issues:
Intersection of the generated surfaces
Potential solutions:
1. increasing the number of tries.
2. Converting .vtk files from intermediate step with a Freesurfer command for topology fix.
### Nipreps documentation (CP)
CP added a document to the Nipreps documentation explaining how to read fMRIprep visual report (pull request: [https://github.com/nipreps/nipreps.github.io/pull/21](https://github.com/nipreps/nipreps.github.io/pull/21)). This document is inspired from an original draft of Saren Seeley and [https://www.frontiersin.org/articles/10.3389/fnimg.2022.1073734/full](https://www.frontiersin.org/articles/10.3389/fnimg.2022.1073734/full).
Problem: The image used to build the website is outdated.
Proposed solution: Set it up to ignore rendering PR.
### fMRIPrep (CM)
fMRIPrep 23.0 milestone: [https://github.com/nipreps/fmriprep/milestone/60](https://github.com/nipreps/fmriprep/milestone/60) Waiting on final merging of #2855.
### fMRIPrep Rodents (EM)
Given questions on Neurostars about fMRIPrep rodents, it becomes necessary to fix the bugs of fMRIPrep Rodents to have a working pipeline.
### Next NiPreps Roundup
* Agenda on HackMD [https://hackmd.io/@oesteban/rJHigdRho/edit](https://hackmd.io/@oesteban/rJHigdRho/edit).
* Governance documents: New pull request with minor fixes [https://github.com/nipreps/GOVERNANCE/pull/19](https://github.com/nipreps/GOVERNANCE/pull/19) (Typos + rule on the number of members in the Steering Committee.)
### MRIQC memory issue (ES)
Synthstrip runs out of memory during MRIQC runs [https://github.com/nipreps/mriqc/issues/1004](https://github.com/nipreps/mriqc/issues/1004)
The problem is caused by the model used in Synthstrip. The first and last convolutional layers generate features maps of ~6GB for images with dimension (192x256x256). When multiple images are processed in parallel MRIQC can exceed the memory limit.
Proposed solutions:
* Limit the number of Synthstrip processes running simultaneously.
* Make sure the configuration produce the smallest possible dimension.