Improved quality of computed tomography substitute derived from magnetic resonance (MR) data by incorporation of spatial information – potential application for MR-only radiotherapy and attenuation correction in positron emission tomography

Figures & data

Figure 1. Spatial distribution of Gaussian components in the mixture. Each coloured shape represents a component. The sharp edges (a) indicate where the Mahalanobis distance from the centre of each component is one. The local densities of the different components are shown in (b) and the corresponding slice in (c).

Figure 2. Mean absolute prediction deviation in the CT substitute with respect to the Hounsfield number in the CT substitute.

Figure 3. CT images (a,f,k,p), CT substitutes derived from a model including spatial variables (b,g,l,q), CT substitutes derived from a model without spatial variables (c,h,m,r) and difference between CT substitute and CT for the model with (d,I,n,s) and without (e,j,o,t) spatial information.