DR-GPT : A large language model for medical report analysis of diabetic retinopathy patients

dc.contributorAalto-yliopistofi
dc.contributorAalto Universityen
dc.contributor.authorJaskari, Joel
dc.contributor.authorSahlsten, Jaakko
dc.contributor.authorSummanen, Paula
dc.contributor.authorMoilanen, Jukka
dc.contributor.authorLehtola, Erika
dc.contributor.authorAho, Marjo
dc.contributor.authorSäpyskä, Elina
dc.contributor.authorHietala, Kustaa
dc.contributor.authorKaski, Kimmo
dc.contributor.departmentDepartment of Computer Scienceen
dc.contributor.groupauthorKaski Kimmo groupen
dc.contributor.organizationDepartment of Computer Science
dc.contributor.organizationUniversity of Helsinki
dc.contributor.organizationCentral Finland Health Care District
dc.date.accessioned2024-10-30T06:37:43Z
dc.date.available2024-10-30T06:37:43Z
dc.date.issued2024-10
dc.descriptionPublisher Copyright: Copyright: © 2024 Jaskari et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
dc.description.abstractDiabetic retinopathy (DR) is a sight-threatening condition caused by diabetes. Screening programmes for DR include eye examinations, where the patient’s fundi are photographed, and the findings, including DR severity, are recorded in the medical report. However, statistical analyses based on DR severity require structured labels that calls for laborious manual annotation process if the report format is unstructured. In this work, we propose a large language model DR-GPT for classification of the DR severity from unstructured medical reports. On a clinical set of medical reports, DR-GPT reaches 0.975 quadratic weighted Cohen’s kappa using truncated Early Treatment Diabetic Retinopathy Study scale. When DR-GPT annotations for unlabeled data are paired with corresponding fundus images, the additional data improves image classifier performance with statistical significance. Our analysis shows that large language models can be applied for unstructured medical report databases to classify diabetic retinopathy with a variety of applications.en
dc.description.versionPeer revieweden
dc.format.extent14
dc.format.mimetypeapplication/pdf
dc.identifier.citationJaskari, J, Sahlsten, J, Summanen, P, Moilanen, J, Lehtola, E, Aho, M, Säpyskä, E, Hietala, K & Kaski, K 2024, ' DR-GPT : A large language model for medical report analysis of diabetic retinopathy patients ', PloS one, vol. 19, no. 10 October, e0297706, pp. 1-14 . https://doi.org/10.1371/journal.pone.0297706en
dc.identifier.doi10.1371/journal.pone.0297706
dc.identifier.issn1932-6203
dc.identifier.otherPURE UUID: f5e4735d-fb25-4a79-8ffd-8de89a6abfe0
dc.identifier.otherPURE ITEMURL: https://research.aalto.fi/en/publications/f5e4735d-fb25-4a79-8ffd-8de89a6abfe0
dc.identifier.otherPURE LINK: http://www.scopus.com/inward/record.url?scp=85206123042&partnerID=8YFLogxK
dc.identifier.otherPURE FILEURL: https://research.aalto.fi/files/162550422/DR-GPT_-_A_large_language_model_for_medical_report_analysis_of_diabetic_retinopathy_patients.pdf
dc.identifier.urihttps://aaltodoc.aalto.fi/handle/123456789/131476
dc.identifier.urnURN:NBN:fi:aalto-202410306991
dc.language.isoenen
dc.publisherPublic Library of Science
dc.relation.ispartofseriesPloS one
dc.relation.ispartofseriesVolume 19, issue 10 October, pp. 1-14
dc.rightsopenAccessen
dc.titleDR-GPT : A large language model for medical report analysis of diabetic retinopathy patientsen
dc.typeA1 Alkuperäisartikkeli tieteellisessä aikakauslehdessäfi
dc.type.versionpublishedVersion
Files