Extending Probabilistic U-Net Using MC-Dropout to Quantify Data and Model Uncertainty
Publication date
2022
Authors
Bhat, IR
Kuijf, Hugo J.
Editors
Crimi, Alessandro
Bakas, Spyridon
Advisors
Supervisors
Document Type
Part of book
Metadata
Show full item recordCollections
License
taverne
Abstract
We extend the Probabilistic U-Net using MC-Dropout to estimate model uncertainty in addition to the data uncertainty in order to improve the overall predictive uncertainty estimate. We use this model on the datasets present in the QUBIQ21 challenge and achieve a mean score of 0.719.
Keywords
Taverne, Theoretical Computer Science, General Computer Science
Citation
Bhat, I & Kuijf, H J 2022, Extending Probabilistic U-Net Using MC-Dropout to Quantify Data and Model Uncertainty. in A Crimi & S Bakas (eds), Brainlesion : Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 7th International Workshop, BrainLes 2021, Held in Conjunction with MICCAI 2021, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12963 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 555-559, 7th International Brain Lesion Workshop, BrainLes 2021, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2021, Virtual, Online, 27/09/21. https://doi.org/10.1007/978-3-031-09002-8_48, conference