% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@ARTICLE{MaierHein:288083,
author = {L. Maier-Hein$^*$ and A. Reinke$^*$ and P. Godau$^*$ and M.
D. Tizabi$^*$ and F. Büttner$^*$ and E. Christodoulou$^*$
and B. Glocker and F. Isensee$^*$ and J. Kleesiek and M.
Kozubek and M. Reyes and M. A. Riegler and M.
Wiesenfarth$^*$ and A. E. Kavur$^*$ and C. H. Sudre and M.
Baumgartner$^*$ and M. Eisenmann$^*$ and D.
Heckmann-Nötzel$^*$ and T. Rädsch$^*$ and L. Acion and M.
Antonelli and T. Arbel and S. Bakas and A. Benis and M. B.
Blaschko and M. J. Cardoso and V. Cheplygina and B. A.
Cimini and G. S. Collins and K. Farahani and L. Ferrer and
A. Galdran and B. van Ginneken and R. Haase and D. A.
Hashimoto and M. M. Hoffman and M. Huisman and P. Jannin and
C. E. Kahn and D. Kainmueller and B. Kainz and A. Karargyris
and A. Karthikesalingam and F. Kofler and A.
Kopp-Schneider$^*$ and A. Kreshuk and T. Kurc and B. A.
Landman and G. Litjens and A. Madani and K. Maier-Hein$^*$
and A. L. Martel and P. Mattson and E. Meijering and B.
Menze and K. G. M. Moons and H. Müller and B. Nichyporuk
and F. Nickel and J. Petersen and N. Rajpoot and N. Rieke
and J. Saez-Rodriguez and C. I. Sánchez and S. Shetty and
M. van Smeden and R. M. Summers and A. A. Taha and A.
Tiulpin and S. A. Tsaftaris and B. Van Calster and G.
Varoquaux and P. Jäger$^*$},
title = {{M}etrics reloaded: recommendations for image analysis
validation.},
journal = {Nature methods},
volume = {21},
number = {2},
issn = {1548-7091},
address = {London [u.a.]},
publisher = {Nature Publishing Group},
reportid = {DKFZ-2024-00337},
pages = {195 - 212},
year = {2024},
note = {#EA:E130#LA:E290#},
abstract = {Increasing evidence shows that flaws in machine learning
(ML) algorithm validation are an underestimated global
problem. In biomedical image analysis, chosen performance
metrics often do not reflect the domain interest, and thus
fail to adequately measure scientific progress and hinder
translation of ML techniques into practice. To overcome
this, we created Metrics Reloaded, a comprehensive framework
guiding researchers in the problem-aware selection of
metrics. Developed by a large international consortium in a
multistage Delphi process, it is based on the novel concept
of a problem fingerprint-a structured representation of the
given problem that captures all aspects that are relevant
for metric selection, from the domain interest to the
properties of the target structure(s), dataset and algorithm
output. On the basis of the problem fingerprint, users are
guided through the process of choosing and applying
appropriate validation metrics while being made aware of
potential pitfalls. Metrics Reloaded targets image analysis
problems that can be interpreted as classification tasks at
image, object or pixel level, namely image-level
classification, object detection, semantic segmentation and
instance segmentation tasks. To improve the user experience,
we implemented the framework in the Metrics Reloaded online
tool. Following the convergence of ML methodology across
application domains, Metrics Reloaded fosters the
convergence of validation methodology. Its applicability is
demonstrated for various biomedical use cases.},
subtyp = {Review Article},
cin = {E130 / FM01 / E230 / C060 / E290},
ddc = {610},
cid = {I:(DE-He78)E130-20160331 / I:(DE-He78)FM01-20160331 /
I:(DE-He78)E230-20160331 / I:(DE-He78)C060-20160331 /
I:(DE-He78)E290-20160331},
pnm = {315 - Bildgebung und Radioonkologie (POF4-315)},
pid = {G:(DE-HGF)POF4-315},
typ = {PUB:(DE-HGF)16},
pubmed = {pmid:38347141},
doi = {10.1038/s41592-023-02151-z},
url = {https://inrepo02.dkfz.de/record/288083},
}