Show simple item record

dc.contributor.authorSaffari, S Ehsan
dc.contributor.authorLöve, Áskell
dc.contributor.authorFredrikson, Mats
dc.contributor.authorSmedby, Örjan
dc.date.accessioned2016-04-14T12:01:45Zen
dc.date.available2016-04-14T12:01:45Zen
dc.date.issued2015en
dc.date.submitted2016en
dc.identifier.citationBMC Med Imaging. 2015, 15:49en
dc.identifier.issn1471-2342en
dc.identifier.pmid26515510en
dc.identifier.doi10.1186/s12880-015-0083-yen
dc.identifier.urihttp://hdl.handle.net/2336/605251en
dc.descriptionTo access publisher's full text version of this article, please click on the hyperlink in Additional Links field or click on the hyperlink at the top of the page marked Files. This article is open access.en
dc.description.abstractFor optimizing and evaluating image quality in medical imaging, one can use visual grading experiments, where observers rate some aspect of image quality on an ordinal scale. To analyze the grading data, several regression methods are available, and this study aimed at empirically comparing such techniques, in particular when including random effects in the models, which is appropriate for observers and patients.
dc.description.abstractData were taken from a previous study where 6 observers graded or ranked in 40 patients the image quality of four imaging protocols, differing in radiation dose and image reconstruction method. The models tested included linear regression, the proportional odds model for ordinal logistic regression, the partial proportional odds model, the stereotype logistic regression model and rank-order logistic regression (for ranking data). In the first two models, random effects as well as fixed effects could be included; in the remaining three, only fixed effects.
dc.description.abstractIn general, the goodness of fit (AIC and McFadden's Pseudo R (2)) showed small differences between the models with fixed effects only. For the mixed-effects models, higher AIC and lower Pseudo R (2) was obtained, which may be related to the different number of parameters in these models. The estimated potential for dose reduction by new image reconstruction methods varied only slightly between models.
dc.description.abstractThe authors suggest that the most suitable approach may be to use ordinal logistic regression, which can handle ordinal data and random effects appropriately.
dc.language.isoenen
dc.publisherBioMed Centralen
dc.relation.urlhttp://dx.doi.org/ 10.1186/s12880-015-0083-yen
dc.relation.urlhttp://www.ncbi.nlm.nih.gov/pmc/articles/PMC4627379/en
dc.rightsArchived with thanks to BMC medical imagingen
dc.subject.meshLogistic Modelsen
dc.subject.meshDiagnostic Imagingen
dc.titleRegression models for analyzing radiological visual grading studies--an empirical comparison.en
dc.typeArticleen
dc.contributor.department[ 1 ] Linkoping Univ, IMH, Dept Med & Hlth Sci, Linkoping, Sweden [ 2 ] Sabzevar Univ Med Sci, Sabzevar, Iran [ 3 ] Lund Univ, Dept Diagnost Radiol, Lund, Sweden [ 4 ] Landspitali Univ Hosp, Dept Radiol, Reykjavik, Iceland [ 5 ] Univ Iceland, Fac Med, Reykjavik, Iceland [ 6 ] Linkoping Univ, Dept Clin & Expt Med, Linkoping, Sweden [ 7 ] KTH Royal Inst Technol, Sch Technol & Hlth, SE-14152 Stockholm, Swedenen
dc.identifier.journalBMC medical imagingen
dc.rights.accessOpen Accessen
refterms.dateFOA2018-09-12T15:53:58Z
html.description.abstractFor optimizing and evaluating image quality in medical imaging, one can use visual grading experiments, where observers rate some aspect of image quality on an ordinal scale. To analyze the grading data, several regression methods are available, and this study aimed at empirically comparing such techniques, in particular when including random effects in the models, which is appropriate for observers and patients.
html.description.abstractData were taken from a previous study where 6 observers graded or ranked in 40 patients the image quality of four imaging protocols, differing in radiation dose and image reconstruction method. The models tested included linear regression, the proportional odds model for ordinal logistic regression, the partial proportional odds model, the stereotype logistic regression model and rank-order logistic regression (for ranking data). In the first two models, random effects as well as fixed effects could be included; in the remaining three, only fixed effects.
html.description.abstractIn general, the goodness of fit (AIC and McFadden's Pseudo R (2)) showed small differences between the models with fixed effects only. For the mixed-effects models, higher AIC and lower Pseudo R (2) was obtained, which may be related to the different number of parameters in these models. The estimated potential for dose reduction by new image reconstruction methods varied only slightly between models.
html.description.abstractThe authors suggest that the most suitable approach may be to use ordinal logistic regression, which can handle ordinal data and random effects appropriately.


Files in this item

Thumbnail
Name:
art_16.pdf
Size:
466.6Kb
Format:
PDF

This item appears in the following Collection(s)

Show simple item record