Question about image acquisition order in HCP

115 views
Skip to first unread message

李佳艺

unread,
Nov 10, 2025, 8:59:21 AM (4 days ago) Nov 10
to HCP-Users
Dear all,
I’ve been going through the HCP pipeline and had a quick question about the order of the scans.
 
In the HCP pipeline process, both structural MRI and functional MRI acquisitions include the BIAS_BC, BIAS_32CH, and fieldmap sequences. In addition, the structural images also include the AFI sequence. For MSMAll, are all of these images mandatory?

According to the reference documentation, the acquisition order is as shown in the attached figure. It looks like before the T1w and T2w scans there’s only one set of BIAS_BC and BIAS_32CH, and after the T1w/T2w scans there’s only one fieldmap collected. But when I check data in unprocessed\3T\T1w_MPR1\ and unprocessed\3T\T2w_SPC1\, I see BIAS_BC, BIAS_32CH, and fieldmap images in both folders. Also, T1w_MPR1\BIAS_BC.nii and T2w_SPC1\BIAS_BC.nii contain different voxel values. I’ve noticed something similar in unprocessed\3T\task\ as well. Could you help me understand how these separate bias/fieldmap images are generated, and why they’d be different?

Thanks a lot for your time!

1.png


Harms, Michael

unread,
Nov 10, 2025, 10:47:03 AM (4 days ago) Nov 10
to hcp-...@humanconnectome.org

 

The BIAS_BC scan in the T1w and T2w folders should be the same.

 

If you think that is not the case, please provide the specific command, and output, that you are using to compare the two.

 

Cheers,

-MH

 

-- 

Michael Harms, Ph.D.

-----------------------------------------------------------

Professor of Psychiatry

Washington University School of Medicine

Department of Psychiatry, Box 8134

660 South Euclid Ave.                        Tel: 314-747-6173

St. Louis, MO  63110                          Email: mha...@wustl.edu

 



--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/181ab1f8-c9e8-49e5-a977-f59820ecaa7dn%40humanconnectome.org.

 


The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

Glasser, Matthew

unread,
Nov 10, 2025, 12:45:03 PM (4 days ago) Nov 10
to hcp-...@humanconnectome.org

We should take a step back here.  What are you trying to do?  Acquire new data?  Reprocess HCP data?

Matt.

Message has been deleted
Message has been deleted
Message has been deleted

李佳艺

unread,
Nov 10, 2025, 11:06:49 PM (3 days ago) Nov 10
to HCP-Users, glas...@wustl.edu
Hi Matt,

We are currently acquiring fMRI data and plan to process it using the HCP pipelines in the future. From the documentation, I noticed that the AFIbias_BC, and bias_32CH sequences mentioned for structural MRI are not directly used in the HCP pipelines.
So, structural MRI preprocessing in the HCP pipelines requires T1‑weightedT2‑weighted, and fieldmap scans?(please correct me if I’m wrong). For resting‑state and task fMRI preprocessing, the bias_BC,bias_32CH, and BOLD_{RL,LR}_SB_SE sequences are all required?

Thank you again for your help!

Best regards,
Jiayi Li

Message has been deleted

Harms, Michael

unread,
Nov 10, 2025, 11:21:16 PM (3 days ago) Nov 10
to 李佳艺, HCP-Users

No, the Bias scans are not required.  And you probably want to acquire the fMRI as AP/PA rather than RL/LR.

 

Check out our HCP Lifespan protocol:

https://www.humanconnectome.org/hcp-protocols

 

Although we now suggest turning on Prescan Normalize on all scans.

 

Cheers,

-MH

 

-- 

Michael Harms, Ph.D.

-----------------------------------------------------------

Professor of Psychiatry

Washington University School of Medicine

Department of Psychiatry, Box 8134

660 South Euclid Ave.                        Tel: 314-747-6173

St. Louis, MO  63110                          Email: mha...@wustl.edu

 

From: 李佳 <ljy106...@gmail.com>
Date: Monday, November 10, 2025 at 10:07 PM
To: HCP-Users <hcp-...@humanconnectome.org>
Cc: "Harms, Michael" <mha...@wustl.edu>
Subject: Re: [hcp-users] Question about image acquisition order in HCP

 

Hi MH,

Previously, when I opened both files in MRIcroGL, I noticed different voxel values at the same location, which made me confused. However, I just double‑checked and confirmed that the two images are actually identical. It might have been an issue with the file download.

We are currently acquiring fMRI data and plan to process it using the HCP pipelines in the future. From the documentation, I noticed that the AFIbias_BC, and bias_32CH sequences mentioned for structural MRI are not directly used in the HCP pipelines.
So, structural MRI preprocessing in the HCP pipelines requires T1‑weightedT2‑weighted, and fieldmap scans?(please correct me if I’m wrong). For resting‑state and task fMRI preprocessing, the bias_BC,bias_32CH, and BOLD_{RL,LR}_SB_SE sequences are all required?

Sorry for the unnecessary trouble, and thank you again for your help!

Best,

Jiayi Li

 

20251110日星期一 UTC+8 23:47:03<mha...@wustl.edu> 写道:

 

The BIAS_BC scan in the T1w and T2w folders should be the same.

 

If you think that is not the case, please provide the specific command, and output, that you are using to compare the two.

 

Cheers,

-MH

 

-- 

Michael Harms, Ph.D.

-----------------------------------------------------------

Professor of Psychiatry

Washington University School of Medicine

Department of Psychiatry, Box 8134

660 South Euclid Ave.                        Tel: 314-747-6173

St. Louis, MO  63110                          Email: mha...@wustl.edu

 

From: 李佳 <ljy106...@gmail.com>
Reply-To: "hcp-...@humanconnectome.org" <hcp-...@humanconnectome.org>
Date: Monday, November 10, 2025 at 7:59 AM
To: HCP-Users <hcp-...@humanconnectome.org>
Subject: [hcp-users] Question about image acquisition order in HCP

 

Dear all,
I’ve been going through the HCP pipeline and had a quick question about the order of the scans.

 
In the HCP pipeline process, both structural MRI and functional MRI acquisitions include the BIAS_BC, BIAS_32CH, and fieldmap sequences. In addition, the structural images also include the AFI sequence. For MSMAll, are all of these images mandatory?


According to the reference documentation, the acquisition order is as shown in the attached figure. It looks like before the T1w and T2w scans there’s only one set of BIAS_BC and BIAS_32CH, and after the T1w/T2w scans there’s only one fieldmap collected. But when I check data in unprocessed\3T\T1w_MPR1\ and unprocessed\3T\T2w_SPC1\, I see BIAS_BC, BIAS_32CH, and fieldmap images in both folders. Also, T1w_MPR1\BIAS_BC.nii and T2w_SPC1\BIAS_BC.nii contain different voxel values. I’ve noticed something similar in unprocessed\3T\task\ as well. Could you help me understand how these separate bias/fieldmap images are generated, and why they’d be different?

Thanks a lot for your time!

 

Image removed by sender.

 

 

--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/181ab1f8-c9e8-49e5-a977-f59820ecaa7dn%40humanconnectome.org.

 


The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

Tim Coalson

unread,
Nov 10, 2025, 11:30:25 PM (3 days ago) Nov 10
to hcp-...@humanconnectome.org, glas...@wustl.edu
bias_BC and bias_32CH scans are not required if you use pre-scan normalize on all scans (this is what we now recommend).  AFI or B1Tx are recommended for transmit bias correction of myelin maps (I believe B1Tx is preferred to AFI; the PseudoTransmit method is not as good, it exists as a fallback only for data that didn't acquire transmit bias scans).

Tim


--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
Message has been deleted

Harms, Michael

unread,
Nov 10, 2025, 11:39:46 PM (3 days ago) Nov 10
to 李佳艺, HCP-Users

Yes, you absolutely need SE-EPI with both polarities, or what we call in that Lifespan protocol "SpinEchoFieldMap" scans.

 

-- 

Michael Harms, Ph.D.

-----------------------------------------------------------

Professor of Psychiatry

Washington University School of Medicine

Department of Psychiatry, Box 8134

660 South Euclid Ave.                        Tel: 314-747-6173

St. Louis, MO  63110                          Email: mha...@wustl.edu

 

From: 李佳 <ljy106...@gmail.com>
Date: Monday, November 10, 2025 at 10:36 PM
To: HCP-Users <hcp-...@humanconnectome.org>
Cc: "Harms, Michael" <mha...@wustl.edu>,
李佳 <ljy106...@gmail.com>
Subject: Re: [hcp-users] Question about image acquisition order in HCP

 

Hi MH,


Thank you so much for your prompt reply — it has been a great help to me!!!

Sorry to bother you again, I noticed in the sequence list that, before the task and resting‑state fMRI acquisitions, there are additional BOLD_LR_SB_SE and BOLD_RL_SB_SE sequences.
The documentation states that these are used for topup distortion correction.
 whether these two sequences are required for fMRI preprocessing in the HCP pipelines?

Thank you very much for your time and guidance.

Best regards,
Best
JiayiLI

 

Error! Filename not specified.

 

 

--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/181ab1f8-c9e8-49e5-a977-f59820ecaa7dn%40humanconnectome.org.

 


The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

 


The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

李佳艺

unread,
Nov 11, 2025, 12:29:41 AM (3 days ago) Nov 11
to HCP-Users, mha...@wustl.edu, 李佳艺
Hi MH,

Thank you so much for your prompt reply — it has been a great help to me!!!

Sorry to bother you again, I noticed in the sequence list that, before the task and resting‑state fMRI acquisitions, there are additional BOLD_LR_SB_SE and BOLD_RL_SB_SE sequences.
The documentation states that these are used for topup distortion correction.
 whether these two sequences are required for fMRI preprocessing in the HCP pipelines?

Thank you very much for your time and guidance.

Best regards,
Best,
JiayiLI

1.png
在2025年11月11日星期二 UTC+8 12:21:16<mha...@wustl.edu> 写道:

李佳艺

unread,
Nov 11, 2025, 12:29:41 AM (3 days ago) Nov 11
to HCP-Users, glas...@wustl.edu

Hi, Matt,

We’re collecting MRI data for a new task, and later on we’d like to run it through the HCP pipeline. So I want to know that for MSMAll, which scan sequences do we absolutely need to acquire?

Thanks so much for your help!

Best,

Jiayi Li

李佳艺

unread,
Nov 11, 2025, 12:33:47 AM (3 days ago) Nov 11
to HCP-Users, mha...@wustl.edu

Hi MH,

Thank you very much for your reply! Previously, when I opened both files in MRIcroGL, I noticed different voxel values at the same location, which made me confused. However, I just double‑checked and confirmed that the two images are actually identical. It might have been an issue with the file download.

We are currently acquiring fMRI data and plan to process it using the HCP pipelines in the future. From the documentation, I noticed that the AFIbias_BC, and bias_32CH sequences mentioned for structural MRI are not directly used in the HCP pipelines.
So, structural MRI preprocessing in the HCP pipelines requires T1‑weightedT2‑weighted, and fieldmap scans?(please correct me if I’m wrong). For resting‑state and task fMRI preprocessing, the bias_BC,bias_32CH, and BOLD_{RL,LR}_SB_SE sequences are all required?

Sorry for the unnecessary trouble, and thank you again for your help!

Best regards,
Jiayi Li


在2025年11月10日星期一 UTC+8 23:47:03<mha...@wustl.edu> 写道:
Message has been deleted
Message has been deleted
Message has been deleted
Message has been deleted

李佳艺

unread,
Nov 11, 2025, 6:04:43 PM (3 days ago) Nov 11
to HCP-Users, tim.c...@gmail.com, glas...@wustl.edu
Hi Tim,

Thank you very much for your reply.

Since the MRI scanner we currently use is different from the one employed in the HCP protocols, we are unable to obtain high‑resolution and high‑quality images for AFI or B1Tx sequences.

In the paper The Minimal Preprocessing Pipelines for the Human Connectome Project, within the section HCP Pipelines: Minimum Acquisition Requirements, AFI is not mentioned as part of the listed minimum requirements.

I would like to confirm whether it is acceptable to omit AFI acquisition when aiming to run MSMAll in the HCP pipelines.


Thank you very much for your time and guidance.

Best regards,
Jiayi Li

Tim Coalson

unread,
Nov 11, 2025, 6:25:31 PM (3 days ago) Nov 11
to 李佳艺, HCP-Users, glas...@wustl.edu
B1Tx/AFI is not required for MSMAll, but it is required for accurate statistical group analysis of myelin maps (a dataset with good T1w and T2w scans but no transmit-bias scan is a partially lost opportunity).  It doesn't need to equal structural resolution.  The PseudoTransmit method uses spin echo fieldmaps (which are the recommended fieldmap type for fMRI anyway) instead of a dedicated transmit-bias scan, but is not as accurate.

The transmit bias pipeline is much newer than the initial publication for the pipelines.

Tim

Reply all
Reply to author
Forward
0 new messages