Mbirn: Enhancement of cardiac MRI analysis

From NAMIC Wiki
Jump to: navigation, search
Home < Enhancement of cardiac MRI analysis

Principal Investigator: Gail Adler, MD, PhD

Goal:
The most common method for studying blood perfusion currently is by using PET scans. This has a number of disadvantages relative to using MR techniques: it is considerably more expensive, MR resolution is about 2 - 3 better and PET involves introducing chemicals into the body. The one big advantage that PET has over MR methods for studying blood perfusion is that there has been extensive validation of PET techniques, which is not yet true of MR techniques. Thus the goal of this study is to demonstrate that using MR is a viable means for studying blood perfusion. This in turn means being able to efficiently process patient datasets so as to accumulate a large number of processed fMRI sequences, which can then be validated. The driving motive is to ultimately standardize the process

Progress so far: Dr Raymond Kwong has assisted Dr Adler in creating myocardial perfusion measurements for 20 subjects from MRI studies, with a further 50 expected by May 2005. Specifically the study is aimed at understanding the effect of an Aldesterone receptor antagonist on myocardial perfusion in patients with diabetes mellitus.

Main immediate technical problem:
After the breath-hold phase is terminated by the patient's resumption of breathing, motion distortion is introduced in the region of interest in the heart, which prevents an accurate representation of the perfusing blood. The way to overcome this to register each time-slice to the initial slice, and thus get an approximation to the undistorted perfusion. Currently this is done manually, a very time consuming process currently being performed by Dr Hilton Joffe working with Dr Adler at BWH.

Current and projected work:
At the Surgical Planning Lab, BWH, a number of projects are under way which it is hoped will result in a significant breakthrough in the current MR methodology for studying blood perfusion, and which will be applied to Dr Adler and her coworkers' project as described above.

  • Work on an improved registration algorithm, which will take much less time, be more accurate and potentially be used automatically or semi-automatically. This work is being done by Gunnar Farneback under the guidance of C.F. Westin and should be completed by the beginning of March. Presumably it could be used standalone at this point. The new algorithm uses the ITK package, which has been adopted by 3D Slicer for computational work, thus making it suitable for inclusion in Slicer.
  • Slightly longer term is the incorporation of the new registration algorithm as the first step in a multi-volume processing pipeline that will enable analysis and visualization of perfusion data within 3D Slicer. Registration will either be available as an independent Slicer module, or will be exposed through the Interval Browser (or Ibrowser), a Slicer module for general multi-volume preview and processing. The Ibrowser, being developed by Wendy Plesniak for the fBIRN project, can currently load and display multi-volume sequences of images in several data formats in Slicer. Ibrowser will be extended to load perfusion sequences, enable their processing through its GUI, permit animated preview and saving. Early work on these features is already underway. Once a sequence has been motion corrected, other existing Slicer tools for defining regions or volumes of interest and for performing various measurements and computation on volumes can be applied to the data. We may discover that additional specialized measurement tools may also be required to do the necessary analysis.