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. |
Year | DOI | Venue |
---|---|---|
2011 | 10.1109/ISBI.2011.5872797 | 2011 8TH IEEE INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING: FROM NANO TO MACRO |
Keywords | Field | DocType |
Liver tumor segmentation, manual segmentation, validation | Pattern recognition,Liver tumor,Medical imaging,Computer science,Segmentation,Image segmentation,Artificial intelligence,Computed tomography,Reference standards,Liver tumor segmentation | Conference |
ISSN | Citations | PageRank |
1945-7928 | 2 | 0.36 |
References | Authors | |
1 | 8 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jan Hendrik Moltz | 1 | 80 | 8.85 |
Stefan Braunewell | 2 | 12 | 2.00 |
J. Ruhaak | 3 | 2 | 0.70 |
Frank Heckel | 4 | 41 | 4.59 |
Sebastiano Barbieri | 5 | 38 | 6.13 |
Lennart Tautz | 6 | 65 | 9.79 |
Horst K. Hahn | 7 | 450 | 72.61 |
Heinz-otto Peitgen | 8 | 1030 | 114.91 |