[SOUND] >> Greetings from Baltimore, Maryland. My name is John McGrady and I'm representing the Johns Hopkins Bloomberg school department of bio statistics and I'm excited to introduce you all to the teaching staff for the course R tutorial for neuro imaging, neuro hacking in R. So we get started, I'd like to have each of you instructors introduce themselves and tell a little bit about themselves. So we'll start with Mr. Muschelli. >> My name's John Muschelli. I'm a 4th year PhD student and I've been doing neuroimaging analysis for the past about six years. >> Miss Sweeney. >> Hi. I'm Elizebeth Sweeney. I'm a 3rd year PhD student. I've been doing research in imaging for about four years. >> Doctor Crainiceanu. I'm Doctor Ciprian Crainiceanu. I'm a professor of biostatistics here in the department and I've been brain imaging for the last five or six years. >> So give everybody a little bit of sense of how you got involved with neuro-imaging. >> So for my master's thesis, I did some work for neuro imaging with respect to FMRI. And I worked in an FMRI lab for about two to three years. And I worked on a stroke clinical trial for another about five. So I got some introduction to imaging analysis and imaging processing there. >> Yeah, so when I was doing my masters here at Hopkins in the biostatistics department, I did a summer internship at the National Institute of Neurological Disease and Stroke. And so I was working in an imaging lab there with Danny Riek. And we were doing imaging and multiple sclerosis, so I really got a taste of how to do work with imaging, about image processing, and ever since then I've been in love with the subject and continue to work on it. >> Yeah, and I started working on multiple sclerosis neuroimaging with Daniel Rike and Brian Coffel and probably my first interaction with nuroimaging was how I crashed my computer with 22 downloads of data [LAUGH] and then trying to make sense of the whole thing and trying to work with my crawlers. >> Excellent. So you guys collectively bring a lot of experience with this. So why did you decide to do this course? >> So, at least for me, after doing a lot of the training I kind of had my feet wet in the field. But Elizabeth and I have been trying to kind of convey what we've learned over the few years to others, and created a short course tutorial. And other segments to try to convey that but the scope is rather limited. >> Yeah and I think we had a one hour short course that we've done and a few statistics conferences, and we wanted to kind of expand on that and really move in to using R for imaging. There's a lot of great packages, doinf neuroimaging in R but we wanted to kind of give a really detailed look at those packages and introduce them to people. So kind of just open up things so that everybody could get involved in this research area that we really love. >> Yeah, my incentive was, I think it took me a long time to learn the basic tools on neuroimaging analysis. It just took a very long time, maybe six months, to even get to a level where I was comfortable with my own knowledge. And once I learned that, I started to think about how can we do this faster? How about two months? How about one week and this is what we tried to do. We just tried to compress the time it takes to learn the basics. >> Excellent, so what are your expectations for students who enroll in this class? >> Well, it's more, the expectations are to try to better understand what neuro imaging's all about. So kind of try to eliminate the fear of complex data analysis and complex data manipulations and plotting. So, just try to eliminate the fear. Try to reduce the time it takes between when you start to have a look at your first image and the time you actually can reasonably work on all things. >> So I believe that a lot of the things we've discussed is that one, the formats of certain data for example, are very different to new users, even users of R for many years and even strong users of R. Working with these images, it wouldn't be as intuitive to do some of the things you would think as a first pass. There are some better ways to do it and we want to kind of get that message out that packages exist that do this well. And here's how to do it. >> Yeah. And I think somebody could start taking this course with almost no background in imaging whatsoever. Where they can't even read an image into R and we can take them from that point to the point where they can do, start to do their own analysis and really start to understand things. >> Excellent, so as we move forward to starting up the course, can you guys give a little insight as to what they're working on these days with respect to neuro imaging? >> So, after working with FMRI for a few years, I've focused a lot on CT imaging, so CAT scans of people who've had hemorrhaging strokes and trying to segment those images to say, this is where the stroke is, this is where the stroke isn't, automatically for hopefully surgical planning and analysis downstream. >> And I've always worked with structural MRI, and mainly with patients who have multiple sclerosis. And I've done segmentations similar to John's. So I've worked on lesion segmentation in these images, and now I'm kind of trying to take the things that I've done with the segmentation and compare those to clinical outcomes and treatments the subjects are on. >> I've been working on a variety of neuro imaging analysis. Typically my research is driven by my students [LAUGH]. They decide what I work on, because they have a lot of energy. They bring a lot of ideas. And I tend to work on MRI, that's how I started, that worked on dynamic contrasting hazards. And that's how I started BC MRI. MRI in general, then John the CT school. We also have a great group here that works on FMRI and other modalities as well. So I'd also like to learn from Brian Coffel and Martin Lundquist about FMRI and all sorts of other imaging, as well. >> Excellent. Well, without further ado, we invite you to start with the course. As you can see, it's going to be quite exciting and on the cutting edge. So we look forward to having you here.
