Navigating prevalence shifts in image analysis algorithm deployment | ||
| dc.conicet.organizacion | University Hospital Heidelberg | |
| dc.conicet.organizacion | Ruprecht Karls Universitat Heidelberg | |
| dc.conicet.organizacion | German Cancer Research Center | |
| dc.conicet.organizacion | German Cancer Research Center | |
| dc.conicet.organizacion | German Cancer Research Center | |
| dc.conicet.organizacion | Consejo Nacional de Investigaciones Científicas y Técnicas. Oficina de Coordinación Administrativa Ciudad Universitaria. Instituto de Investigación en Ciencias de la Computación. Universidad de Buenos Aires. Facultad de Ciencias Exactas y Naturales. Instituto de Investigación en Ciencias de la Computación | |
| dc.conicet.organizacion | German Cancer Research Center | |
| dc.conicet.organizacion | German Cancer Research Center | |
| dc.contributor.author | Godau, Patrick | |
| dc.contributor.author | Kalinowski, Piotr | |
| dc.contributor.author | Christodoulou, Evangelia | |
| dc.contributor.author | Reinke, Annika | |
| dc.contributor.author | Tizabi, Minu | |
| dc.contributor.author | Ferrer, Luciana | |
| dc.contributor.author | Jäger, Paul | |
| dc.contributor.author | Maier Hein, Lena | |
| dc.date.issued | 2025-05 | |
| dc.description.abstract | Domain gaps are significant obstacles to the clinical implementation of machine learning (ML) solutions for medical image analysis. Although current research emphasizes new training methods and network architectures, the specific impact of prevalence shifts on algorithms in real-world applications is often overlooked. Differences in class frequencies between development and deployment data are crucial, particularly for the widespread adoption of artificial intelligence (AI), as disease prevalence can vary greatly across different times and locations. Our contribution is threefold. Based on a diverse set of 30 medical classification tasks (1) we demonstrate that lack of prevalence shift handling can have severe consequences on the quality of calibration, decision threshold, and performance assessment. Furthermore, (2) we show that prevalences can be accurately and reliably estimated in a data-driven manner. Finally, (3) we propose a new workflow for prevalence-aware image classification that uses estimated deployment prevalences to adjust a trained classifier to a new environment, without requiring additional annotated deployment data. Comprehensive experiments indicate that our proposed approach could contribute to generating better classifier decisions and more reliable performance estimates compared to current practice. | es |
| dc.description.fil | Fil: Godau, Patrick. University Hospital Heidelberg; Alemania | |
| dc.description.fil | Fil: Kalinowski, Piotr. Ruprecht Karls Universitat Heidelberg; Alemania | |
| dc.description.fil | Fil: Christodoulou, Evangelia. German Cancer Research Center; Alemania | |
| dc.description.fil | Fil: Reinke, Annika. German Cancer Research Center; Alemania | |
| dc.description.fil | Fil: Tizabi, Minu. German Cancer Research Center; Alemania | |
| dc.description.fil | Fil: Ferrer, Luciana. Consejo Nacional de Investigaciones Científicas y Técnicas. Oficina de Coordinación Administrativa Ciudad Universitaria. Instituto de Investigación en Ciencias de la Computación. Universidad de Buenos Aires. Facultad de Ciencias Exactas y Naturales. Instituto de Investigación en Ciencias de la Computación; Argentina | |
| dc.description.fil | Fil: Jäger, Paul. German Cancer Research Center; Alemania | |
| dc.description.fil | Fil: Maier Hein, Lena. German Cancer Research Center; Alemania | |
| dc.format | application/pdf | es |
| dc.identifier.citation | Godau, Patrick; Kalinowski, Piotr; Christodoulou, Evangelia; Reinke, Annika; Tizabi, Minu; et al.; Navigating prevalence shifts in image analysis algorithm deployment; Elsevier Science; Medical Image Analysis; 102; 5-2025; 1-14 | es |
| dc.identifier.institucional | CONICET | |
| dc.identifier.issn | 1361-8415 | es |
| dc.identifier.repo | CONICET Digital | |
| dc.identifier.uri | http://hdl.handle.net/11336/290818 | |
| dc.journal.ciudad | Amsterdam | es |
| dc.journal.pagination | 1-14 | es |
| dc.journal.pais | Países Bajos | es |
| dc.journal.title | Medical Image Analysis | es |
| dc.journal.volume | 102 | es |
| dc.language.iso | eng | es |
| dc.publisher | Elsevier Science | es |
| dc.relation.alternativeid | info:eu-repo/semantics/altIdentifier/url/https://linkinghub.elsevier.com/retrieve/pii/S1361841525000520 | |
| dc.relation.alternativeid | info:eu-repo/semantics/altIdentifier/doi/http://dx.doi.org/10.1016/j.media.2025.103504 | |
| dc.rights | info:eu-repo/semantics/openAccess | * |
| dc.rights.uri | https://creativecommons.org/licenses/by/2.5/ar/ | * |
| dc.subject | Prevalence shift | es |
| dc.subject | Medical image classification | es |
| dc.subject | Generalization | es |
| dc.subject | Domain gap | es |
| dc.subject.classification | Ciencias de la Información y Bioinformática | es |
| dc.subject.classification | Ciencias de la Computación e Información | es |
| dc.subject.classification | CIENCIAS NATURALES Y EXACTAS | es |
| dc.title | Navigating prevalence shifts in image analysis algorithm deployment | es |
| dc.type | info:eu-repo/semantics/article | |
| dc.type | info:ar-repo/semantics/artículo | |
| dc.type | info:eu-repo/semantics/publishedVersion | |
Archivos
Bloque original
Cargando...
- Nombre:
- CONICET_Digital_Nro.29bddf0c-528b-4ca0-a330-6e2e3afa6fb3_B.pdf
- Tamaño:
- 3.1 MB
- Formato:
- Adobe Portable Document Format