DIPY workshop question
Hi, I’m interested in the DIPY workshop but have a couple questions to decide if its right for me. I’ve been using DTI in my research but primarily through software packages such as FSL, Diffusion toolkit, DTI studio. I haven’t tried using DIPY or python for anything yet, but I’m comfortable with unix, FSL, and several other command line based neuroimaging tools. But I’m primarily interested in learning more advanced reconstruction models including modeling free water and IVIM 1. Does the workshop focus primarily on using and troubleshooting DIPY’s available tools or is there an emphasis on the underlying theory? 2. Will my lack of baseline knowledge of DIPY and python be a hindrance? Should I work through some tutorials prior to the workshop, would that be enough to prepare me? Thanks, Kyle Kern, MD Vascular Neurology Clinical Research Fellow NIH, NINDS, Stroke Branch
Hi Kyle, Thanks for your email. I'll tell you my opinion and then others can chime in. My answers inline below. On Fri, Jan 10, 2020 at 11:29 AM Kern, Kyle (NIH/NINDS) [E] via Dipy < dipy@python.org> wrote:
I think that the workshop aims to balance three main things: 1. Solid knowledge of diffusion MRI and its applications. There are sessions on various aspects of the analysis, including a lot of the underlying theory. I am not ashamed to admit that last year's workshop finally filled in several holes that I had in my understanding of dMRI analysis, with some of the best people in the field giving clear and authoritative explanations. 2. New cutting edge technologies in diffusion MRI and its analysis. These sessions focus on new techniques for acquisition and analysis of dMRI data. We are fortunate to have two of the world's foremost experts on dMRI acquisition teaching sessions on their own work: Jennifer McNab (Stanford) and Susie Huang (MGH). We are also going to have the chance to hear about new machine learning techniques (e.g., deep learning) and how these are applied to dMRI data. This is a preview of the next decade of research and applications in the field. 3. Hands-on instruction on how to implement these things in DIPY. More about this below...
To answer this question: In the third category (applications in DIPY), we do show some Python code, and it's not a bad idea to have some sense of how Python works (there are several good online resources for learning about Python and its use in science. I would recommend Jake Vanderplas's book: https://jakevdp.github.io/PythonDataScienceHandbook/). BUT: most of the instruction relies on our workflows, which are command-line interfaces just like the ones you use to run FSL. That is, there is no Python programming involved at all, and instead, you type commands in the command line to run these. In addition, there will be many opportunities to work closely with the development team, to implement analysis pipelines with the data that you intend to analyze. We recommend that attendees bring data with them and we will devote time to individual work on data (with guidance from the instructors and from the DIPY development team). If you don't have data yet, we also have several example datasets of different kinds that you can work with. For more details, I suggest you take a look at the tentative workshop schedule that is now available on the workshop website: https://workshop.dipy.org/ Finally, another benefit of the workshop is that you will have an opportunity to interact with other researchers in the field in person. This is a great opportunity to expand your professional network in the field, hear what other people are working on and present your own work (the workshop includes poster sessions and power pitch sessions). I hope that answers your questions, but please do let us know if you have other questions. Cheers, Ariel
Hi Kyle, Thanks for your email. I'll tell you my opinion and then others can chime in. My answers inline below. On Fri, Jan 10, 2020 at 11:29 AM Kern, Kyle (NIH/NINDS) [E] via Dipy < dipy@python.org> wrote:
I think that the workshop aims to balance three main things: 1. Solid knowledge of diffusion MRI and its applications. There are sessions on various aspects of the analysis, including a lot of the underlying theory. I am not ashamed to admit that last year's workshop finally filled in several holes that I had in my understanding of dMRI analysis, with some of the best people in the field giving clear and authoritative explanations. 2. New cutting edge technologies in diffusion MRI and its analysis. These sessions focus on new techniques for acquisition and analysis of dMRI data. We are fortunate to have two of the world's foremost experts on dMRI acquisition teaching sessions on their own work: Jennifer McNab (Stanford) and Susie Huang (MGH). We are also going to have the chance to hear about new machine learning techniques (e.g., deep learning) and how these are applied to dMRI data. This is a preview of the next decade of research and applications in the field. 3. Hands-on instruction on how to implement these things in DIPY. More about this below...
To answer this question: In the third category (applications in DIPY), we do show some Python code, and it's not a bad idea to have some sense of how Python works (there are several good online resources for learning about Python and its use in science. I would recommend Jake Vanderplas's book: https://jakevdp.github.io/PythonDataScienceHandbook/). BUT: most of the instruction relies on our workflows, which are command-line interfaces just like the ones you use to run FSL. That is, there is no Python programming involved at all, and instead, you type commands in the command line to run these. In addition, there will be many opportunities to work closely with the development team, to implement analysis pipelines with the data that you intend to analyze. We recommend that attendees bring data with them and we will devote time to individual work on data (with guidance from the instructors and from the DIPY development team). If you don't have data yet, we also have several example datasets of different kinds that you can work with. For more details, I suggest you take a look at the tentative workshop schedule that is now available on the workshop website: https://workshop.dipy.org/ Finally, another benefit of the workshop is that you will have an opportunity to interact with other researchers in the field in person. This is a great opportunity to expand your professional network in the field, hear what other people are working on and present your own work (the workshop includes poster sessions and power pitch sessions). I hope that answers your questions, but please do let us know if you have other questions. Cheers, Ariel
participants (2)
-
Ariel Rokem -
Kern, Kyle (NIH/NINDS) [E]