Journal Article (Review Article)

DKFZ-2024-00337

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Metrics reloaded: recommendations for image analysis validation.

Maier-Hein, L. (First author)DKFZ* ; Reinke, A. (First author)DKFZ* ; Godau, P.DKFZ* ; Tizabi, M. D.DKFZ* ; Büttner, F.DKFZ* ; Christodoulou, E.DKFZ* ; Glocker, B. ; Isensee, F.DKFZ* ; Kleesiek, J. ; Kozubek, M. ; Reyes, M. ; Riegler, M. A. ; Wiesenfarth, M.DKFZ* ; Kavur, A. E.DKFZ* ; Sudre, C. H. ; Baumgartner, M.DKFZ* ; Eisenmann, M.DKFZ* ; Heckmann-Nötzel, D.DKFZ* ; Rädsch, T.DKFZ* ; Acion, L. ; Antonelli, M. ; Arbel, T. ; Bakas, S. ; Benis, A. ; Blaschko, M. B. ; Cardoso, M. J. ; Cheplygina, V. ; Cimini, B. A. ; Collins, G. S. ; Farahani, K. ; Ferrer, L. ; Galdran, A. ; van Ginneken, B. ; Haase, R. ; Hashimoto, D. A. ; Hoffman, M. M. ; Huisman, M. ; Jannin, P. ; Kahn, C. E. ; Kainmueller, D. ; Kainz, B. ; Karargyris, A. ; Karthikesalingam, A. ; Kofler, F. ; Kopp-Schneider, A.DKFZ* ; Kreshuk, A. ; Kurc, T. ; Landman, B. A. ; Litjens, G. ; Madani, A. ; Maier-Hein, K.DKFZ* ; Martel, A. L. ; Mattson, P. ; Meijering, E. ; Menze, B. ; Moons, K. G. M. ; Müller, H. ; Nichyporuk, B. ; Nickel, F. ; Petersen, J. ; Rajpoot, N. ; Rieke, N. ; Saez-Rodriguez, J. ; Sánchez, C. I. ; Shetty, S. ; van Smeden, M. ; Summers, R. M. ; Taha, A. A. ; Tiulpin, A. ; Tsaftaris, S. A. ; Van Calster, B. ; Varoquaux, G. ; Jäger, P. (Last author)DKFZ*

2024
Nature Publishing Group London [u.a.]

This record in other databases:  

Please use a persistent id in citations: doi:10.1038/s41592-023-02151-z

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.

Note: #EA:E130#LA:E290#

---

Contributing Institute(s):

1. E130 Intelligente Medizinische Systeme (E130)
2. DKTK Koordinierungsstelle Frankfurt (FM01)
3. E230 Medizinische Bildverarbeitung (E230)
4. C060 Biostatistik (C060)
5. NWG Interaktives maschinelles Lernen (E290)

Research Program(s):

1. 315 - Bildgebung und Radioonkologie (POF4-315) (POF4-315)

Appears in the scientific report 2024

---

Database coverage:
; BIOSIS Previews ; Biological Abstracts ; Clarivate Analytics Master Journal List ; Current Contents - Life Sciences ; DEAL Nature ; Ebsco Academic Search ; Essential Science Indicators ; IF >= 40 ; JCR ; Nationallizenz ; SCOPUS ; Science Citation Index Expanded ; Web of Science Core Collection

---