Paper Info
Title
On the Ambiguity of Registration Uncertainty.
Abstract
Estimating the uncertainty in image registration is an area of current research that is aimed at providing information that will enable surgeons to assess the operative risk based on registered image data and the estimated registration uncertainty. If they receive inaccurately calculated registration uncertainty and misplace confidence in the alignment solutions, severe consequences may result. For probabilistic image registration (PIR), most research quantifies the registration uncertainty using summary statistics of the transformation distributions. In this paper, we study a rarely examined topic: whether those summary statistics of the transformation distribution truly represent the registration uncertainty. Using concrete examples, we show that there are two types of uncertainties: the transformation uncertainty, Ut, and label uncertainty Ul. Ut indicates the doubt concerning transformation parameters and can be estimated by conventional uncertainty measures, while Ul is strongly linked to the goal of registration. Further, we show that using Ut to quantify Ul is inappropriate and can be misleading. In addition, we present some potentially critical findings regarding PIR.
Year
Venue
Field
2018
arXiv: Computer Vision and Pattern Recognition
Operative risk,Computer science,Artificial intelligence,Summary statistics,Probabilistic logic,Ambiguity,Machine learning,Image registration
DocType
Volume
Citations 
Journal
abs/1803.05266
0
PageRank 
References 
Authors
0.34
10
12
Name
Order
Citations
PageRank
Jie Luo1125.58
Sarah F. Frisken233.43
Karteek Popuri3598.80
Dana Cobzas420722.19
Frank Preiswerk5777.16
Matthew Toews624720.60
Miaomiao Zhang713226.12
Hongyi Ding801.01
Polina Golland91690114.38
Alexandra J Golby109613.93
Masashi Sugiyama113353264.24
William M. Wells III125267833.10