Missing value in dHCP pipeline report

1

Dear all,

This is Hailong Li, a research associate in Imaging research center, Cincinnati Children’s Hospital Medical Center, US.

We are using the dHCP structural pipeline to process our neonatal T2 images. First, I would like to express our appreciation to dHCP structural team for the excellent neonatal structural pipeline.

We have a quick question here:
After running the pipeline and obtaining pipeline measures (e.g., volume, thickness), we found that there are some missing values in the report.

Specifically, we found that (GI - Insula right & GI - Insula left) measures contained the most part of missing values marked as ‘-’.

We thought that the insula was not correctly segmented, but actually, other metrics (volumes, thickness) of insula were not missing and clearly provided in the report.

Could someone here please help us understand the issue?

Thanks

Hailong

2

Hi,

I got the same problem as you did, but my missing value is 'GI - Medial and inferior temporal gyri posterior part right ’ which occurs on 2 subjects of 91.

In addition, do you mind telling me your results of sulcal depth and curvature, just their numerical range? Because my results of sulcal depth and curvature are quite strange.

Thank you in advance!
Xinyi

3

Can you clarify what you mean with “report”? Which file(s) are you referring to?

Thanks, Daniel

4

Hello, Daniel
We are using the measures in: \reports\pipeline_all_measures.csv.

Hailong

5

Hello, Xinyi
I took a very rough estimation on our neonates cohort.
sulc depth is -200 to +100
curvature is -2 to +3.

Hailong

6

We will get back to you soon.

7

Hi,
Thank you for the prompt response! My results are similar with yours, but seem unreasonable. I think the whole brain sulcal depth should be around 2~3 as some papers reported. Meanwhile the whole brain curvature should be 0.2 or so.

Xinyi

8

Dear Hailong,

do you run the pipeline yourself or do you use the files which are provided with the second data release?

Thanks, Daniel

9

Hi, Daniel
We used the dHCP structural pipeline docker to process our own neonatal T2 MRI, collected around 40 weeks.
Thanks
Hailong