Quantifying Uncertainty in Neural Network Segmentation

By |2022-07-26T08:36:27-04:00June 8th, 2022|

Dr Jones and Sair, along with Collaborators in the School of Public Health and the Division of Neuroradiology have designed a method to quantify two types of uncertainty, aleatoric and epistemic, directly from a neural network segmentation algorithm.  Details can be found here: Direct quantification of epistemic and aleatoric uncertainty in 3D U-net segmentation https://doi.org/10.1117/1.JMI.9.3.034002