Student Research Projects and Skills

Most MRI research is focused towards one or more clinical applications, and involves some of the technical developments.
Thus I have split the topics into clinical and technical:

Clinical Applications

  • Musculoskeletal Imaging: Osteoarthritis affects almost all of the population at some stage of life. MRI is the most accurate non-invasive method of diagnosing cartilage lesions. Non-invasive imaging would reduce the costs assocated with arthroscopy, and allow earlier treatment of osteoarthritis. More importantly, as treatments of arthritis evolve, MRI can be used to assess the response of cartilage to treatments. Cartilage imaging demands efficient imaging with high contrast between cartilage and surrounding tissues. We are exploring the application of new imaging methods to cartilage imaging. Much of this work is already in the clinical arena, as musculoskeletal imaging already accounts for as many as 70% of body MRI exams.
  • Breast Imaging: Breast cancer is one of the most commonly occurring cancers, which, with early diagnosis is also one of the most treatable. Current screening techniques, primarily X-ray mammography lack sensitivity to early stage tumors. MR screening is sensitive, but not yet specific enough to justify the increased cost. New contrast mechanisms, combined with rapid, high resolution imaging will be essential in this area where MRI has huge potential for reducing mortality.
  • Cardiovascular Imaging: Cardiovascular disease is the leading cause of death in developed countries. Non-invasive imaging of the cardiovascular system continues to be a hot research area, as early diagnosis of heart disease allows quicker and simpler treatment. Rapid imaging methods are very useful in cardiac imaging, as they are less susceptible to motion artifacts. Rapid imaging can also be used for contrast-enhanced perfusion.
  • Abdominal Imaging: MRI of non-cardiac areas of the abdomen is still limited by motion artifacts and spatial resolution when compared to X-ray computed tomography (CT). Implementation of current and new techniques for motion correction and more rapid scanning continues to be important for most areas of Abdominal MRI.

Technical Development

  • RF Pulse Design: There are many areas where new excitation pulses can be tailored to specific applications. There are a few areas where completely new design frameworks would be useful as well.
  • Reconstruction Algorithms: This is a very broad area that includes non-Cartesian image reconstruction (such as gridding), various corrections for off-resonance, gradient imperfections and undersampling, as well as retrospective contrast mechanisms (for example fat/water separation), and finally segmentation and analysis tools.
  • Rapid Imaging: Rapid imaging methods include multiple spin echo sequences, rapid gradient echo, spoiled gradient echo and balanced SSFP. I have worked predominently on balanced SSFP, which is one of the most efficient of these methods, but which sill has numerous technical limitations. All of these methods will play some role in most of the above clinical applications.

Programming an MRI scanner is a skill that takes C programmers background 8-12 months to learn. This is the most useful skill that a student in MRI pulse sequence development can learn and continue to develop.


3D femoral cartilage images.
(2:30 Scan time)


3D Lower Leg Vasculature
(75 second Scan time)

Prerequisites

Skills

The following are skills that I expect students or postdocs to have, or to develop quickly.
If you do not have one of these skills, don't try to hide the fact (it will be obvious!), but please talk to me about it!
You will have to continue to independently maintain these skills by working with others and reading books or websites.
  1. Communication Skills: The ability to write scientific reports is a basic skill required for research. You can develop this by practice, and reading. Closely linked is the ability to give oral presentations. Never pass up the opportunity to present your work to an audience!
  2. Unix/Linux: Some familiarity with Unix/Linux is important for working within the university environment, as well as the GE MRI environment.
  3. Matlab: Programming in Matlab is virtually essential for research in MRI, as it allows you to quickly test algorithms before coding in C/C++. You can learn Matlab in a few weeks, and in many EE classes.
  4. C/C++ : The ability to write clear, commented program code in C and/or C++ is vital for writing usable software for sequence design, or image reconstruction. Skills in debugging programs are twice as important as the ability to write them!
  5. Office Apps: The industry standard for software includes Microsoft Word, Excel and Powerpoint. You should be functional in using all of these; the more you know, the better life will be.

Courses

The following are courses that you should take (or equivalent) prior to working in medical imaging. If I say it is required, it really is!

Additional Points

When requesting positions in my research group, strong preference will be given to students who: I will usually only accept students who have fit at least one of these categories

Application Process

At Stanford, PhD students are admitted to an academic department. Before being assigned a thesis advisor, students must receive formal acceptance to a department. For work in MRI, students typically are admitted to Electrical Engineering for the MS or PhD program.

Please read the General Stanford Admission web site and/or the Electrical Engineering admissions site. I will be pleased to answer questions relating to research areas and research projects, but you should contact the appropriate departments with admissions questions.


General Advice / Contacting

This is general advice if you are seeking a research group or an advisor. If these are obvious, you are probably in good shape! I will NOT respond to inquiries about research that do not show genuine interest.