Collaborative training of medical artificial intelligence models with non-uniform labels
Due to the rapid advancements in recent years, medical image analysis is largely dominated by deep learning (DL). However, building powerful and robust DL models requires training with large multi-party datasets. While multiple stakeholders have provided publicly available datasets, the ways in whic...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article (Journal) |
| Language: | English |
| Published: |
13 April 2023
|
| In: |
Scientific reports
Year: 2023, Volume: 13, Pages: 1-9 |
| ISSN: | 2045-2322 |
| DOI: | 10.1038/s41598-023-33303-y |
| Online Access: | Verlag, kostenfrei, Volltext: https://doi.org/10.1038/s41598-023-33303-y Verlag, kostenfrei, Volltext: https://www.nature.com/articles/s41598-023-33303-y 
|
| Author Notes: | Soroosh Tayebi Arasteh, Peter Isfort, Marwin Saehn, Gustav Mueller-Franzes, Firas Khader, Jakob Nikolas Kather, Christiane Kuhl, Sven Nebelung & Daniel Truhn |
MARC
| LEADER | 00000caa a2200000 c 4500 | ||
|---|---|---|---|
| 001 | 1854328131 | ||
| 003 | DE-627 | ||
| 005 | 20240307063332.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 230804s2023 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1038/s41598-023-33303-y |2 doi | |
| 035 | |a (DE-627)1854328131 | ||
| 035 | |a (DE-599)KXP1854328131 | ||
| 035 | |a (OCoLC)1425213400 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rda | ||
| 041 | |a eng | ||
| 084 | |a 33 |2 sdnb | ||
| 100 | 1 | |a Tayebi Arasteh, Soroosh |e VerfasserIn |0 (DE-588)1298471109 |0 (DE-627)1854454927 |4 aut | |
| 245 | 1 | 0 | |a Collaborative training of medical artificial intelligence models with non-uniform labels |c Soroosh Tayebi Arasteh, Peter Isfort, Marwin Saehn, Gustav Mueller-Franzes, Firas Khader, Jakob Nikolas Kather, Christiane Kuhl, Sven Nebelung & Daniel Truhn |
| 264 | 1 | |c 13 April 2023 | |
| 300 | |a 9 | ||
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Gesehen am 04.08.2023 | ||
| 520 | |a Due to the rapid advancements in recent years, medical image analysis is largely dominated by deep learning (DL). However, building powerful and robust DL models requires training with large multi-party datasets. While multiple stakeholders have provided publicly available datasets, the ways in which these data are labeled vary widely. For Instance, an institution might provide a dataset of chest radiographs containing labels denoting the presence of pneumonia, while another institution might have a focus on determining the presence of metastases in the lung. Training a single AI model utilizing all these data is not feasible with conventional federated learning (FL). This prompts us to propose an extension to the widespread FL process, namely flexible federated learning (FFL) for collaborative training on such data. Using 695,000 chest radiographs from five institutions from across the globe—each with differing labels—we demonstrate that having heterogeneously labeled datasets, FFL-based training leads to significant performance increase compared to conventional FL training, where only the uniformly annotated images are utilized. We believe that our proposed algorithm could accelerate the process of bringing collaborative training methods from research and simulation phase to the real-world applications in healthcare. | ||
| 650 | 4 | |a Biomedical engineering | |
| 650 | 4 | |a Computer science | |
| 700 | 1 | |a Isfort, Peter |e VerfasserIn |4 aut | |
| 700 | 1 | |a Saehn, Marwin |e VerfasserIn |4 aut | |
| 700 | 1 | |a Mueller-Franzes, Gustav |e VerfasserIn |4 aut | |
| 700 | 1 | |a Khader, Firas |e VerfasserIn |4 aut | |
| 700 | 1 | |a Kather, Jakob Nikolas |d 1989- |e VerfasserIn |0 (DE-588)1064064914 |0 (DE-627)812897587 |0 (DE-576)423589091 |4 aut | |
| 700 | 1 | |a Kuhl, Christiane |e VerfasserIn |4 aut | |
| 700 | 1 | |a Nebelung, Sven |e VerfasserIn |4 aut | |
| 700 | 1 | |a Truhn, Daniel |e VerfasserIn |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Scientific reports |d [London] : Springer Nature, 2011 |g 13(2023), Artikel-ID 6046, Seite 1-9 |h Online-Ressource |w (DE-627)663366712 |w (DE-600)2615211-3 |w (DE-576)346641179 |x 2045-2322 |7 nnas |a Collaborative training of medical artificial intelligence models with non-uniform labels |
| 773 | 1 | 8 | |g volume:13 |g year:2023 |g elocationid:6046 |g pages:1-9 |g extent:9 |a Collaborative training of medical artificial intelligence models with non-uniform labels |
| 856 | 4 | 0 | |u https://doi.org/10.1038/s41598-023-33303-y |x Verlag |x Resolving-System |z kostenfrei |3 Volltext |
| 856 | 4 | 0 | |u https://www.nature.com/articles/s41598-023-33303-y |x Verlag |z kostenfrei |3 Volltext |
| 951 | |a AR | ||
| 992 | |a 20230804 | ||
| 993 | |a Article | ||
| 994 | |a 2023 | ||
| 998 | |g 1064064914 |a Kather, Jakob Nikolas |m 1064064914:Kather, Jakob Nikolas |d 910000 |d 910100 |e 910000PK1064064914 |e 910100PK1064064914 |k 0/910000/ |k 1/910000/910100/ |p 5 | ||
| 999 | |a KXP-PPN1854328131 |e 4363103760 | ||
| BIB | |a Y | ||
| SER | |a journal | ||
| JSO | |a {"id":{"eki":["1854328131"],"doi":["10.1038/s41598-023-33303-y"]},"title":[{"title_sort":"Collaborative training of medical artificial intelligence models with non-uniform labels","title":"Collaborative training of medical artificial intelligence models with non-uniform labels"}],"recId":"1854328131","note":["Gesehen am 04.08.2023"],"name":{"displayForm":["Soroosh Tayebi Arasteh, Peter Isfort, Marwin Saehn, Gustav Mueller-Franzes, Firas Khader, Jakob Nikolas Kather, Christiane Kuhl, Sven Nebelung & Daniel Truhn"]},"person":[{"family":"Tayebi Arasteh","roleDisplay":"VerfasserIn","given":"Soroosh","role":"aut","display":"Tayebi Arasteh, Soroosh"},{"family":"Isfort","roleDisplay":"VerfasserIn","display":"Isfort, Peter","role":"aut","given":"Peter"},{"roleDisplay":"VerfasserIn","family":"Saehn","role":"aut","display":"Saehn, Marwin","given":"Marwin"},{"given":"Gustav","display":"Mueller-Franzes, Gustav","role":"aut","family":"Mueller-Franzes","roleDisplay":"VerfasserIn"},{"display":"Khader, Firas","role":"aut","given":"Firas","roleDisplay":"VerfasserIn","family":"Khader"},{"given":"Jakob Nikolas","role":"aut","display":"Kather, Jakob Nikolas","family":"Kather","roleDisplay":"VerfasserIn"},{"roleDisplay":"VerfasserIn","family":"Kuhl","given":"Christiane","role":"aut","display":"Kuhl, Christiane"},{"given":"Sven","display":"Nebelung, Sven","role":"aut","family":"Nebelung","roleDisplay":"VerfasserIn"},{"role":"aut","display":"Truhn, Daniel","given":"Daniel","roleDisplay":"VerfasserIn","family":"Truhn"}],"physDesc":[{"extent":"9 S."}],"relHost":[{"part":{"year":"2023","extent":"9","pages":"1-9","text":"13(2023), Artikel-ID 6046, Seite 1-9","volume":"13"},"physDesc":[{"extent":"Online-Ressource"}],"language":["eng"],"type":{"bibl":"periodical","media":"Online-Ressource"},"origin":[{"publisher":"Springer Nature ; Nature Publishing Group","dateIssuedDisp":"2011-","publisherPlace":"[London] ; London","dateIssuedKey":"2011"}],"title":[{"title":"Scientific reports","title_sort":"Scientific reports"}],"id":{"issn":["2045-2322"],"zdb":["2615211-3"],"eki":["663366712"]},"note":["Gesehen am 12.07.24"],"disp":"Collaborative training of medical artificial intelligence models with non-uniform labelsScientific reports","pubHistory":["1, article number 1 (2011)-"],"recId":"663366712"}],"language":["eng"],"type":{"bibl":"article-journal","media":"Online-Ressource"},"origin":[{"dateIssuedDisp":"13 April 2023","dateIssuedKey":"2023"}]} | ||
| SRT | |a TAYEBIARASCOLLABORAT1320 | ||