University of Cambridge > > Isaac Newton Institute Seminar Series > Machine Learned Priors for Nonsmooth Conductivities in D-bar Reconstructions of 2D EIT Data

Machine Learned Priors for Nonsmooth Conductivities in D-bar Reconstructions of 2D EIT Data

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact nobody.

RNTW02 - Rich and non-linear tomography in medical imaging, materials and non destructive testing

Recent developments in D-bar reconstruction methods for 2D EIT have established a methodology for the inclusion of spatial priors, which have been shown to provide improved quality and stability of images. In the context of medical imaging, these techniques begin with prior estimates of organ boundaries within the plane of the electrodes, to which optimized conductivity guesses are assigned. In previous works, the methodology for approximating organ boundaries has involved manually extracting boundaries from prior medical scans, which may not be readily available in practice. This protocol is also highly labor intensive, and has the potential to introduce human bias. Furthermore, in previous works, some of the sharpness provided by the introduction of priors was lost due to a mathematical need for smoothing of the conductivity distribution. In this presentation, we address these problems via (1) a method for the automated selection of boundaries via machine learning techniques, and (2) use of an alternative mathematical formulation which eliminates the need for smoothing. We present results from numerically simulated thoracic phantoms on circular domains.

This talk is part of the Isaac Newton Institute Seminar Series series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.


© 2006-2023, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity