Analysis of Variability in Manual Liver Tumor Delineation in CT Scans

Jan Hendrik Moltz, Stefan Braunewell, Jan Rühaak, Frank Heckel, Sebastiano Barbieri, Lennart Tautz, Horst Karl Hahn, Heinz-Otto Peitgen

Abstract

Manual delineations by experts are often used as reference standards for validating segmentation algorithms, although it is well known that they always show some degree of variability. Our goal is to estimate the effects of using a limited number of expert segmentations. Given ten manual delineations of 13 liver tumors, we analyzed the volume error made by randomly selecting subsets of the ten segmentations compared to the complete set. We found that when using just one segmentation the expected error was 17% with a maximum of 35%. This means that it is questionable whether a comparison with a single reference allows a reliable validation. When three segmentations are chosen, the error is halved, so this might be a reasonable compromise between accuracy and viability of evaluation studies.
Original languageEnglish
Title of host publication2011 IEEE International Symposium on Biomedical Imaging: From Nano to Macro
Number of pages4
Place of PublicationChicago, Illinois, USA
PublisherIEEE
Publication date01.03.2011
Pages1974-1977
ISBN (Print)978-1-4244-4127-3
ISBN (Electronic)978-1-4244-4128-0
DOIs
Publication statusPublished - 01.03.2011
Event2011 IEEE International Symposium on Biomedical Imaging: From Nano to Macro - Chicago, United States
Duration: 30.03.201102.04.2011

Fingerprint

Dive into the research topics of 'Analysis of Variability in Manual Liver Tumor Delineation in CT Scans'. Together they form a unique fingerprint.

Cite this