Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms

This paper investigates the limits of foundation models in medical image segmentation, mainly focusing on SAM by Meta. While previous research demonstrated SAM’s potential for cost-efficient segmentation, this study explores its performance enhancement through integration with prompt enhancement opt...

Descripción completa

Detalles Bibliográficos
Autores: Gutiérrez, Juan D., Lozano García, Nuria, Delgado Muñoz, Emilio, Rubio Largo, Álvaro, Rodríguez Echeverría, Roberto
Tipo de recurso: artículo
Fecha de publicación:2026
País:España
Institución:Universidad de Santiago de Compostela (USC)
Repositorio:Minerva. Repositorio Institucional de la Universidad de Santiago de Compostela
Idioma:inglés
OAI Identifier:oai:minerva.usc.gal:10347/46141
Acceso en línea:https://hdl.handle.net/10347/46141
Access Level:acceso abierto
Palabra clave:Deep Learning
Foundation Models
Genetic Algorithms
Image Segmentation
Medical Imaging
Zero-Shot Learning
id ES_34ffa4e8514f82d0d04d2707af11e022
oai_identifier_str oai:minerva.usc.gal:10347/46141
network_acronym_str ES
network_name_str España
repository_id_str
spelling Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithmsGutiérrez, Juan D.Lozano García, NuriaDelgado Muñoz, EmilioRubio Largo, ÁlvaroRodríguez Echeverría, RobertoDeep LearningFoundation ModelsGenetic AlgorithmsImage SegmentationMedical ImagingZero-Shot LearningThis paper investigates the limits of foundation models in medical image segmentation, mainly focusing on SAM by Meta. While previous research demonstrated SAM’s potential for cost-efficient segmentation, this study explores its performance enhancement through integration with prompt enhancement optimization and genetic algorithms, aiming to minimize user input further. As a proof of concept, we apply this novel approach to lung segmentation tasks using public axial lung CT scans, frontal chest X-ray datasets, and spleen MRIs. Our findings reveal that the genetic algorithm optimization significantly improves SAM’s segmentation accuracy, bringing it closer to the state-of-the-art performance achieved by specifically trained models. In particular, when compared with our previous approach, this technique reaches a 94.85 % Jaccard Index (+3.77 delta) and a 97.17 % Dice Score (+2.50 delta) for lung CT scans, a 93.39 % Jaccard Index (+5.95 delta) and a 96.57 %Dice Score (+3.38 delta) for chest X-rays, and a 91.00 % Jaccard Index (+6.51 delta) and a 95.07 % Dice Score (+4.12 delta) for spleen MRIs. Notably, this improvement is achieved without retraining or modifying SAM’s architecture. However, our analysis also identifies an inherent limitation in this optimization approach, revealing a performance ceiling that cannot be surpassed despite further genetic algorithm iterations. The implications of these findings emphasize the potential of combining foundation models with non-intrusive optimization techniques for cost-effective and accessible medical image segmentation. While dataset-related limitations may affect generalizability, validating the approach across broader clinical scenarios remains essential. Future work should explore applications to additional organs, diverse datasets, and the integration of expert-in-the-loop strategies to enhance clinical utilityUNIRUniversidade de Santiago de Compostela. Departamento de Electrónica e Computación20262026-02-2420262026-02-24journal articlehttp://purl.org/coar/resource_type/c_6501VoRhttp://purl.org/coar/version/c_970fb48d4fbd8a85info:eu-repo/semantics/articleapplication/pdfhttps://hdl.handle.net/10347/46141reponame:Minerva. Repositorio Institucional de la Universidad de Santiago de Compostelainstname:Universidad de Santiago de Compostela (USC)InglésengAgencia Estatal de Investigación http://dx.doi.org/10.13039/501100011033 Plan Estatal de Investigación Científica y Técnica y de Innovación 2021-2023 PID2022-137275NA-I00 METAHEURISTICAS PARALELAS MULTIARQUITECTURA Y ENERGETICAMENTE EFICIENTES PARA BIOINFORMATICAopen accesshttp://purl.org/coar/access_right/c_abf2Authors transfer copyright of the article to the publisher UNIR and agree that the article will be distributed under the terms of the Creative Commons Attribution 3.0 unported Licensehttp://creativecommons.org/licenses/by/4.0/info:eu-repo/semantics/openAccessoai:minerva.usc.gal:10347/461412026-06-15T12:47:27Z
dc.title.none.fl_str_mv Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms
title Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms
spellingShingle Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms
Gutiérrez, Juan D.
Deep Learning
Foundation Models
Genetic Algorithms
Image Segmentation
Medical Imaging
Zero-Shot Learning
title_short Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms
title_full Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms
title_fullStr Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms
title_full_unstemmed Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms
title_sort Exploring the limits of foundation models in medical image segmentation: a case study with SAM and genetic algorithms
dc.creator.none.fl_str_mv Gutiérrez, Juan D.
Lozano García, Nuria
Delgado Muñoz, Emilio
Rubio Largo, Álvaro
Rodríguez Echeverría, Roberto
author Gutiérrez, Juan D.
author_facet Gutiérrez, Juan D.
Lozano García, Nuria
Delgado Muñoz, Emilio
Rubio Largo, Álvaro
Rodríguez Echeverría, Roberto
author_role author
author2 Lozano García, Nuria
Delgado Muñoz, Emilio
Rubio Largo, Álvaro
Rodríguez Echeverría, Roberto
author2_role author
author
author
author
dc.contributor.none.fl_str_mv Universidade de Santiago de Compostela. Departamento de Electrónica e Computación
dc.subject.none.fl_str_mv Deep Learning
Foundation Models
Genetic Algorithms
Image Segmentation
Medical Imaging
Zero-Shot Learning
topic Deep Learning
Foundation Models
Genetic Algorithms
Image Segmentation
Medical Imaging
Zero-Shot Learning
description This paper investigates the limits of foundation models in medical image segmentation, mainly focusing on SAM by Meta. While previous research demonstrated SAM’s potential for cost-efficient segmentation, this study explores its performance enhancement through integration with prompt enhancement optimization and genetic algorithms, aiming to minimize user input further. As a proof of concept, we apply this novel approach to lung segmentation tasks using public axial lung CT scans, frontal chest X-ray datasets, and spleen MRIs. Our findings reveal that the genetic algorithm optimization significantly improves SAM’s segmentation accuracy, bringing it closer to the state-of-the-art performance achieved by specifically trained models. In particular, when compared with our previous approach, this technique reaches a 94.85 % Jaccard Index (+3.77 delta) and a 97.17 % Dice Score (+2.50 delta) for lung CT scans, a 93.39 % Jaccard Index (+5.95 delta) and a 96.57 %Dice Score (+3.38 delta) for chest X-rays, and a 91.00 % Jaccard Index (+6.51 delta) and a 95.07 % Dice Score (+4.12 delta) for spleen MRIs. Notably, this improvement is achieved without retraining or modifying SAM’s architecture. However, our analysis also identifies an inherent limitation in this optimization approach, revealing a performance ceiling that cannot be surpassed despite further genetic algorithm iterations. The implications of these findings emphasize the potential of combining foundation models with non-intrusive optimization techniques for cost-effective and accessible medical image segmentation. While dataset-related limitations may affect generalizability, validating the approach across broader clinical scenarios remains essential. Future work should explore applications to additional organs, diverse datasets, and the integration of expert-in-the-loop strategies to enhance clinical utility
publishDate 2026
dc.date.none.fl_str_mv 2026
2026-02-24
2026
2026-02-24
dc.type.none.fl_str_mv journal article
http://purl.org/coar/resource_type/c_6501
VoR
http://purl.org/coar/version/c_970fb48d4fbd8a85
dc.type.openaire.fl_str_mv info:eu-repo/semantics/article
format article
dc.identifier.none.fl_str_mv https://hdl.handle.net/10347/46141
url https://hdl.handle.net/10347/46141
dc.language.none.fl_str_mv Inglés
eng
language_invalid_str_mv Inglés
language eng
dc.relation.none.fl_str_mv Agencia Estatal de Investigación http://dx.doi.org/10.13039/501100011033 Plan Estatal de Investigación Científica y Técnica y de Innovación 2021-2023 PID2022-137275NA-I00 METAHEURISTICAS PARALELAS MULTIARQUITECTURA Y ENERGETICAMENTE EFICIENTES PARA BIOINFORMATICA
dc.rights.none.fl_str_mv open access
http://purl.org/coar/access_right/c_abf2
http://creativecommons.org/licenses/by/4.0/
dc.rights.openaire.fl_str_mv info:eu-repo/semantics/openAccess
rights_invalid_str_mv open access
http://purl.org/coar/access_right/c_abf2
http://creativecommons.org/licenses/by/4.0/
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv UNIR
publisher.none.fl_str_mv UNIR
dc.source.none.fl_str_mv reponame:Minerva. Repositorio Institucional de la Universidad de Santiago de Compostela
instname:Universidad de Santiago de Compostela (USC)
instname_str Universidad de Santiago de Compostela (USC)
reponame_str Minerva. Repositorio Institucional de la Universidad de Santiago de Compostela
collection Minerva. Repositorio Institucional de la Universidad de Santiago de Compostela
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869405853068558336
score 15,812429

Similares en LA Referencia