JB's Blog

I was just shown this link for looking up journal impact factors on the ISI Web of Knowledge site:

http://admin-apps.isiknowledge.com/JCR/JCR?RQ=HOME

Notes and thoughts on neuroscience journal impact factors:
1) Neuron is a nose ahead of Nature Neuroscience, 14.17 vs. 14.16.
2) Why do I not come across more references to papers in Annual Reviews Neuroscience, which is at the top of the list?
3) Neuroimage and Human Brain Mapping, the two imaging journals I'm most familiar with, are tied at 5.5.
4) My favorite reviews are always in Nature Reviews Neuroscience, which makes sense - they're second on the list, impact factor 25.94.

I started trying out Mendeley yesterday. It's a .pdf organizer similar to Papers, but free. I need something like this in order to organize all my .pdf's locally since Zotero is embedded in Firefox. I'm going to try and import my Zotero library into Mendeley and see how that goes. If these two programs can integrate well, I'll be much happier about the state of my papers. I'm still trying to figure out how best to integrate the two (and whether I actually need both). The article lookup via Google Scholar in Mendeley doesn't seem as good as the search that Papers does. You can use Mendeley on PCs and Macs where Papers only works on Macs. So the jury is still out here.
http://www.mendeley.com/

Yesterday Fritjof Helmchen from the University of Zurich spoke at the Joint Seminars in Neuroscience. He talked about using two photon microscopy with calcium dyes to image hundreds of neurons within a large 3-dimensional volume with millisecond level resolution. His talk was very good, this technology is awesome and improving. He was imaging neurons and glia within mouse barrel cortex in vivo in a 200um cubic volume. This is getting up towards the low end of high-resolution fMRI voxel size, which of course interests me as an imager. In the near future we'll be able to use technology like this to see how the activity of a population of cells relates to the BOLD response in a much more complete way. Here's his website.
http://www.hifo.uzh.ch/research/neurophysiology/helmchen/interest.html

Rubniov and Sporns have a new article in Neuroimage describing network graph theory metrics in detail. A good reference to have.
http://dx.doi.org/10.1016/j.neuroimage.2009.10.003

I'm getting my talk ready for Society for Neuroscience 2009 in Chicago. My working title is: Altered Functional Connectivity in Healthy Older Subjects at Genetic Risk for Alzheimer's Disease

Here's the session info:

Session Type: Nanosymposium
Session Number: 212
Session Title: Alzheimer's Disease: Imaging and Biomarkers
Date and TIme: Sunday Oct 18, 2009 1:00PM - 3:30PM
Location: Room S102

It should be a nice quick talk. Come check it out.

The new Scientific American cover features cognitive enhancers. The article was decent, though not really bringing anything new to the table. The article referenced within, from Henry Greely et al. in Nature last year, is where the real interesting discussion takes place.

http://dx.doi.org/10.1038/456702a

I don't know how much neuroimaging work is being done studying cognitive enhancers. This makes me think there needs to be more. Both one-shot experimental studies and longitudinal studies with subjects taking enhancers and controls needs to be done. We'll learn several things from these experiments: what brain regions are modulated? what are the short term effects (e.g. enhanced attention, larger working memory capacity, verbal fluidity, expansion of creativity) and what are the neural correlates? what are the long term effects and side effects?

On another note, I picked up Zack Lynch's new book The Neuro Revolution: How Brain Science is Changing Our World this weekend. It's a good read so far. His long chapter about how neuroimaging will change law, lie detection, pleading insanity, etc. was thought provoking. The message is what's really interesting. Initially I was apprehensive about the magnitude of his claims but I'm becoming a bit more convinced.

This new paper in Neuron from Hannula and Ranganath looks very interesting. They found that hippocampal activity predicted the correct memory in a relational memory task even when explicit conscious recall failed.

http://dx.doi.org/10.1016/j.neuron.2009.08.025

I like the quote from Kumaran and Wagner's preview of the article:
"an alternative view of hippocampal mnemonic function, namely that the hippocampus is critical for relational memory—i.e., memory for the relations between the individual elements of an experience—and contributes to performance irrespective of whether the participant is aware or unaware that relational knowledge has been retrieved"

Laurel and I are leading an FSL Workshop for at least the next quarter. We want to help new people learn the main FSL tools and progress to more advanced topics (ROI analysis, functional connectivity, complex GLM modeling) later in the course. We also hope to bring together a lot of the different labs using FSL at UCLA so we can pool our knowledge and share tools. We are meeting once a week, currently Tuesdays from 10:30-11:30. We've made a website for the Workshop on the UCLA Center for Cognitive Neuroscience wiki that includes our powerpoint slides and useful links. We're also planning to set up a mailing list where we can discuss problems. Here's the website:

http://www.ccn.ucla.edu/wiki/index.php/FSL_Group

If you're interested in coming to the course email me and I'll add you to the list.

I'm participating in the UCLA NITP advance fMRI summer course right now, as student and helper. There have been some really good talks on MR physics (Mark Cohen), the origin and interpretability of the BOLD signal (Rick Buxton), and experimental design (Susan Bookheimer and Russ Poldrack). For any interested in watching, there's a live webcast:

http://www.brainmapping.org/Live.php

Went to the OHBM conference in San Francisco last week, my first. My takeaways, in no particular order:

- Networks are the thing.
- Resting state fMRI is far more popular than I realized.
- There's more overlap in individual representations of concepts (e.g. words) than I previously thought. See Tom Mitchell's talk.
- Low profile, high resolution MRI goggles are going to be big.
- 7T is happening. Resistance is futile.
- Dopamine is so hot right now.
- Don't forget about neurons. Don't ever forget about neurons. In imaging, I think we tend to over-rely on simple ideas, like that all cortex is the same, that all connections are the same, that the brain can simply be divided into "gray matter", "white matter", and "CSF." The brain is a variable, heterogenous, dynamic, and chaotic electrochemical gelatin.

I recently discovered TrackVis, a program for performing and displaying tractography from DTI, DSI, Q-ball, and HARDI data.

http://www.trackvis.org

It's the most intuitive and stable of the tractography programs I've tried. Benefits:

1) It's coupled with Diffusion Toolkit, a good program for calculating tensors from diffusion weighted images. This program has shed some light on otherwise dark areas for me like the idiosyncracies of gradient vectors from different manufacturers (Siemens, GE, etc.). It can also adjust vectors for you if the axis of acquisition was different from the scanner default.
2) The tractography runs like a breeeze, doesn't crash my new iMac, and offers four different propagation algorithms.
3) TrackVis (the actual tract viewing program) is highly intuitive. Generating ROIs for running tractography is a piece of cake, either by loading in a nifti image and setting a threshold for voxel values as an ROI, or using a sphere as an ROI. Then you just create tract from ROI and it renders it right quick.
4) It gives stats on the tracts including mean FA and mean length.

My one complaint is that the online documentation is very brief. Then again, I haven't taken the time to watch the tutorials yet so I shouldn't speak too soon. In any case, after two hours of playing around I had mostly figured out how to use the program. I am anxious to try this with some DSI or HARDI data, my sense is that the major tracts are the only ones that are faithfully represented in DTI data.

Here's an image I produced:



Here's the reference:
Wang R, Wedeen VJ (2007) Diffusion Toolkit and Trackvis. Berlin: Proc Intl Soc Mag Reson Med.

This blog is dedicated to her.

Such a Kevin Kelly ripoff. Anyway:

Use Zotero. It's awesome. Manage your references with it, it's an add-on to Firefox. Anytime you view a page that is an abstract/research paper, click the Z icon and the reference is stored to your library. Then get the Microsoft Word add-on and when you want to add a citation, just click the Zotero menu and add citation, with any standard formatting you choose. Then add a bibliography and boom. I haven't messed with Endnote and something tells me my life is enriched by not doing so.

http://www.zotero.org/

Papers: A very nice (Mac only) tool for managing all your .pdf's. I have used so many different PDF naming schemes over the years. Due to drift and general disorganization, my PDF collection is scattered, redundant, and a mess. Papers helps.

http://mekentosj.com/papers/

And finally - Google Analytics is my new best friend. Sign up, it's free, post a piece of javascript onto each page of your website and then get bukustats on who visits, how they found you, what they look at, where they're from, etc. So cool.

http://www.google.com/analytics/

Two new Ph.D.'s in three days from the Bookheimer Lab. Well done!

1.
Nicola Filippini, Bradley J. MacIntosh, Morgan G. Hough, Guy M. Goodwin, Giovanni B. Frisoni, Stephen M. Smith, Paul M. Matthews, Christian F. Beckmann, and Clare E. Mackay
Distinct patterns of brain activity in young carriers of the APOE-e4 allele
PNAS 2009 106: 7209-7214.
http://dx.doi.org/10.1073/pnas.0811879106

2.

Complex brain networks: graph theoretical analysis of structural and functional systems

Ed Bullmore & Olaf Sporns

Nature Reviews Neuroscience 10, 312 (2009)
http://dx.doi.org/10.1038/nrn2575

Hello World.