IMPROVED VISUAL APPEARANCE OF MR-CAT IMAGES BY ADDING CT-LIKE NOISE
Publication Date: 2016-Feb-18
The IP.com Prior Art Database
Page 01 of 1
Improved visual appearance of MR-CAT images by adding CT-like noise
The work relates to generation of CT-like data from MR data. Notably, from(multi- dimensional) MR information tissue classes are identified and subsequently corresponding CT attenuation data are computed. This results in an MR-calculated attenuation image (MR- CAT image).
In this work CT-like noise is added to the MR-CAT image. Generation of the MR- CAT data involves segmentation which causes image regions of constant grey values. This causes artificial looking MR-CAT images. Adding CT-like noise leads to a more realistic appearance of the MR-CAT image that is easier for radiation oncologists to trust.
This work relates to the field of MR-only based radiotherapy planning. The procedure requires estimation of an electron density (ED) map from the MR image, which can be achieved in the workflow by producing a CT-like image from the MR image, called the MR- CAT image. Since MR image values do not correspond uniquely to ED map values, this estimation process requires advanced image processing and a good visualization of the MR- CAT image in order to assess potential errors. This ID discloses a method that modifies the visual appearance of an MR-CAT image by adding noise to it, in such a way that it looks more like the CT image a clinician is accustomed to evaluate. This eases the assessment work of the clinician and increases the trust in preceding image processing. The ED map continues...