Fig. 1 | Scientific Reports

Fig. 1

From: Complex-valued neural networks to speed-up MR thermometry during hyperthermia using Fourier PD and PDUNet

Fig. 1

An example of non-invasive PRF shift MR thermometry to monitor and control temperature during clinical hyperthermia (HT). The initial temperature map at time step t0 requires two MR images, which are calculated by voxel-wise subtracting the second phase image from the first reference phase image. This reference image is subtracted from the phase images of further acquired MR images, which are taken every 10 minutes during HT therapy. In the initial high-precision magnitude MR image, the temperature map is shown as a color overlay (blue: relative temperature decrease; green: constant temperature; red: relative temperature rise).

Back to article page