Abstract #1584

MR to pseudo CT conversion: Combining Deep-Learning and Analytical Image Processing

Florian Wiesinger¹, Sandeep Kaushik², Mathias Engström³, Pauline Hinault^4,5, Andrew Leynes⁶, Mikael Bylund⁷, David Gensanne⁸, Tufve Nyholm⁷, Peder Larson⁶, and Cristina Cozzini¹

¹GE Healthcare, Munich, Germany, ²GE Global Research, Bangalore, India, ³GE Healthcare, Stockholm, Sweden, ⁴LITIS, Rouen, France, ⁵GE Healthcare, Paris, France, ⁶Department of Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States, ⁷Umeå University, Umea, Sweden, ⁸Centre Henri Becquerel, Rouen, France

Here we present an improved method for ZTE to pseudo CT conversion by combining an analytical signal model (i.e. ZTE to CT signal scaling) with Connected Component Analysis (CCA) and Deep Learning (DL) based air vs. bone discrimination. The method is demonstrated for the two main anatomical regions (head&neck and pelvis) and the two main field strengths (1.5T and 3T) of interest.

This abstract and the presentation materials are available to members only; a login is required.

Join Here