Overview

HFM is a tool set designed to enable one to examine the subregions that comprise the medial temporal lobe (MTL) by combining high-resolution structural and functional magnetic resonance imaging in humans. Utilizing cortical unfolding techniques, one can visualize the activity across the different cellular compartments of the MTL and explore the roles they play at different stages of memory formation and retrieval..

My goals for this project are twofold:
1) Offer to scientists interested in studying the detailed circuitry of the medial temporal lobe the means to study it with fMRI
2) Pool resources across institutions to advance these tools

Hopefully, by putting my work out there, others interested will help contribute to this work.

Everything here is AS-IS, WITHOUT SUPPORT. If you plan to pursue this project, let me be very clear. It assumes sound knowledge of fMRI, MATLAB, UNIX, and image processing. Furthermore, it requires persistence. If something does not work, it will be up to you to debug it.

If you get stuck, I created an email/posting list at the sourceforge.net website (http://lists.sourceforge.net/mailman/listinfo/mtl-unfolding-help). I strongly recommend signing up to this list and searching or submitting to it when you have questions. That way, all the users of HFM can help each other out. I will read posts when I have time, but I often will not have the time. The hope is that knowledge will disseminate among the users so everyone pitching together can make things work.

I do ask that I not be sent me direct email for support, and requested that people read/email the list for that reason. I receive all of the email on the list. Feel free to email me directly if we have talked in person, if you have major project direction questions, or if you are interested in collaborations and/or consulting.

This toolset evolved from tools created by the lab of Brian Wandell at Stanford (whom I thank gratefully!) When I was coding my tools to unfold the hippocampus a couple of years ago, the version of the unfolding tool, mrUnfold, was 3.0, and the version of mrLoadRet, the visualization tool, was 1.0. My tools are tightly linked to these versions. A newer version of the unfolding algorithm based on different theory has since been developed by Wandell's lab, but I have yet to integrate these newer tools into my code. Nevertheless, this only fuels the motivation this open source project: to facilitate the integrated development and advancement of tools that will enable us to explore the details of the medial temporal lobe.

The immediate challenges are as follows:
1) File formats - right now, all the code works with GE scanner output, from an old scanner at that, using an outdated ANMR format for the EPI images. I do have some scattered routines that can handle analyze format, but it will be a clear need for this project for the code to accommodate other formats. The task is actually not extraordinary: the motion correction routines already use analyze format with the AIR package. Exact details would be:

Changing the motion correction to accept analyze data as input
Changing the initial data importing step to accept analyze EPI
Changing the structural importing step to accept analyze
Confirming that importing motion corrected data works correctly
Confirming silly things like R/L, A/P, S/I do not get screwed up in the process.

2) Integration with newer versions of mrUnfold and mrLoadRet

3) Implementing B-field-distortion corrections before importing EPI data.

Long term challenges:
1) Automated segmentation - gray, white, and CSF.
2) Automated delineation of subregions.
3) More precise anatomical localization of subregion boundaries to cover regions like the presubiculum

References:
Please read the following 3 papers for full background:
1. Zeineh MM, Engel SA, Bookheimer, SY. Application of Cortical Unfolding Techniques to Functional MRI of the Human Hippocampal Region. NeuroImage 11(6):668-83, 2000
2. Zeineh MM, Engel SA, Thompson PM, Bookheimer SY. Unfolding the Human Hippocampus with High Resolution Structural and Functional MRI. The Anatomical Record: The New Anatomist 265:111-120, 2001.
3. Zeineh MM, Engel SA, Thompson PM, Bookheimer SY. Dynamics of the hippocampus during encoding and retrieval of face-name pairs. Science. 2003 Jan 24;299(5606):577-80 .

Also, you absolutely need to have the following 3 references before you could hope to delineate the MTL anatomy. Doing anything without these references is a guarantee for disaster.
1. Amaral, D. G., and Insausti, R. 1990. Hippocampal formation. In The Human Nervous System (G. Praxinos, Ed.), pp. 711–755. Academic Press, San Diego.
2. Duvernoy, H. M. 1998. The Human Hippocampus: Functional Anatomy, Vascularization, and Serial Sections with MRI. Springer, Berlin.
3. Insausti, R., Juottonen, K., Soininen, H., Insausti, A., Partanen, K., Vainio, P., Laakso, M., and Pitkanen, A. 1998. MR volumetric analysis of the human entorhinal, perirhinal, and temporopolar cortices. Am. J. Neuroradiol. 19: 659–671.

Thanks to all those who have helped put this project together, in particular Susan Bookheimer, Steve Engel, Russ Poldrack, and Paul Rodriguez.

Michael Zeineh, M.D.-Ph.D.

Back to main