Karsten Sommer1, Jochen Keupp1, Christophe Schuelke1, Oliver Lips1, and Tim Nielsen1
1Philips Research, Hamburg, Germany
1Philips Research, Hamburg, Germany
An uncertainty metric based on test-time augmentation is explored for 3D
motion artifact correction. Analysis on test data revealed that this metric
could accurately predict global error levels but failed in certain cases to
detect local “hallucinations”.
Example results from the synthetic test dataset
where the uncertainty metric Δy accurately predicted the validity of the
corrections: For the input image shown in the top row, the network produces a
faithful reconstruction of the ground truth image, which is accurately
predicted by the low uncertainty metric. For the center and bottom row,
substantial deviations from the ground truth can be observed (marked by
arrows), which is accurately predicted by the relatively high uncertainty
metric (SD=standard deviation).
Example results on the synthetic test dataset
where the uncertainty metric Δy did not accurately
predict the validity of the corrections. In particular, multiple local
hallucinations can be observed (indicated by red arrows), although the TTA
metric values are relatively low (Δy=0.3).