Roberto Souza1,2, Jordan McEwen3, Carissa Chung3, and Ashley D. Harris2,4
1Electrical and Computer Engineering, University of Calgary, Calgary, AB, Canada, 2Hotchkiss Brain Institute, Calgary, AB, Canada, 3Biomedical Engineering, University of Calgary, Calgary, AB, Canada, 4Radiology, University of Calgary, Calgary, AB, Canada
1Electrical and Computer Engineering, University of Calgary, Calgary, AB, Canada, 2Hotchkiss Brain Institute, Calgary, AB, Canada, 3Biomedical Engineering, University of Calgary, Calgary, AB, Canada, 4Radiology, University of Calgary, Calgary, AB, Canada
A
simple flat Convolutional Neural Network architecture investigated in the
context of reducing J-difference edited MRS acquisition times by factors of 4,
8, and 16 shows great potential reduce acquisition times for edited-MRS.
Figure 1.
Architecture
of the CNN model used in the experiments. The model processes the ON and OFF
transients separately and subtracts the results to obtain the final J-difference
edited spectrum.
Figure 4. Sample spectra reconstructed
by the different CNN models for the different samples in the test set.