Fleiss’ Kappa scores for QC inter-rater variability experiments.

Fleiss’ Kappa scores for QC inter-rater variability experiments.

<p>Fleiss’ Kappa agreement scores for QC pipelines across four raters suggest high inter-rater variability. Different levels of agreement as presented in Landis and Koch20 are shaded in with labels to the right of the plot. SLANT-TICV, TractSeg CC4, and TractSeg CSTleft are QCed with “almost p...

Full description

Saved in:
Bibliographic Details
Main Author: Michael E. Kim (21956727) (author)
Other Authors: Chenyu Gao (9501623) (author), Nancy R. Newlin (21956730) (author), Gaurav Rudravaram (21956733) (author), Aravind R. Krishnan (21956736) (author), Karthik Ramadass (13949238) (author), Praitayini Kanakaraj (9193371) (author), Kurt G. Schilling (9193353) (author), Blake E. Dewey (19066193) (author), David A. Bennett (229710) (author), Sid O’Bryant (6449120) (author), Robert C. Barber (21956739) (author), Derek Archer (21956742) (author), Timothy J. Hohman (8356254) (author), Shunxing Bao (21956745) (author), Zhiyuan Li (205620) (author), Bennett A. Landman (7623857) (author), Nazirah Mohd Khairi (21956748) (author)
Published: 2025
Subjects:
Biophysics
Cell Biology
Science Policy
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
quantitative outlier detection
thorough quality control
rater variability experiment
several processing pipelines
proposed method satisfies
low time cost
robust visual qc
scalable quality control
div >< p
automated qc method
following design criteria
data processing pipelines
quality data
scalable manner
design criteria
comparison experiment
time spent
time consuming
qc pipeline
yet necessary
slight differences
quick visualization
preprocessed data
erroneous conclusions
easily shared
also provide
algorithm failure
3 .)
2 .)
1 .)
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:<p>Fleiss’ Kappa agreement scores for QC pipelines across four raters suggest high inter-rater variability. Different levels of agreement as presented in Landis and Koch20 are shaded in with labels to the right of the plot. SLANT-TICV, TractSeg CC4, and TractSeg CSTleft are QCed with “almost perfect” agreement, whereas TractSeg AF right, TractSeg SLFIleft, TractSeg TPOSTCright, and Lung CT Harmonization have “substantial” agreement. PreQual is the only pipeline QCed with “fair” agreement.</p>

Similar Items