Diabētiskās retinopātijas skrīnings optometrista praksē
Loading...
Date
Authors
Advisor
Journal Title
Journal ISSN
Volume Title
Publisher
Latvijas Universitāte
Language
lav
Abstract
Maģistra darbs uzrakstīts latviešu valodā uz 46 lapām, satur 21 attēlu, 12 tabulas, 2 pielikumus un 105 atsauces uz literatūras avotiem. Darba mērķis: novērtēt optometristu un automātiskās skrīninga programmatūras prasmes atpazīt diabētiskās retinopātijas pazīmes. Pētījuma dalībnieki: pētījumā piedalījās 2 oftalmologi, 14 optometristi un automatizētā skrīninga programmatūra AI Avenue. Metode: 196 tīklenes attēliem tika veikts skrīninga tests atbilstoši starptautiskās diabētiskās retinopātijas klasifikācijai. Rezultāti: Rekomendēto testa standartu – jutība un specifiskums ≥ 80% sasniedza oftalmologs un 2 optometristi. AI Avenue sasniedza vienīgi rekomendēto testa specifiskumu. Secinājumi: AI Avenue ļoti labi atpazīst gadījumus, kad nav saslimšana, savukārt optometristiem ir tendence kļūdaini pozitīvi diagnosticēt slimību. Atslēgas vārdi: cukura diabēts, diabētiskā retinopātija, diabētiskā makulas tūska, automatizēta skrīninga programmatūra.
Master thesis is written in Latvian. It contains 46 pages, 21 images, 12 tables, 105 references and 2 attachments. Purpose: To evaluate the skills of optometrists and automatic diabetic retinopathy analysis software to recognize the signs of diabetic retinopathy. Participants: The study included 17 participants – 2 ophthalmologists, 14 optometrists and 1 automatic diabetic retinopathy screening software AI Avenue. Methods: The grading according International Diabetic Retinopathy Classification of 196 retinal images provided by participants. Results: The recommended test standard – sensitivity and specificity are more or equal to 80% was reached by 3 participants – ophthalmologist and 2 optometrists. AI Avenue reached only recommended test specificity ≥ 80.00%. Conclusions: AI Avenue recognizes cases of non-disease very well, while optometrists tend to misdiagnose the grade of disease. Key words: diabetes mellitus, diabetic retinopathy, diabetic macular oedema, automated detection of diabetic retinopathy.
Master thesis is written in Latvian. It contains 46 pages, 21 images, 12 tables, 105 references and 2 attachments. Purpose: To evaluate the skills of optometrists and automatic diabetic retinopathy analysis software to recognize the signs of diabetic retinopathy. Participants: The study included 17 participants – 2 ophthalmologists, 14 optometrists and 1 automatic diabetic retinopathy screening software AI Avenue. Methods: The grading according International Diabetic Retinopathy Classification of 196 retinal images provided by participants. Results: The recommended test standard – sensitivity and specificity are more or equal to 80% was reached by 3 participants – ophthalmologist and 2 optometrists. AI Avenue reached only recommended test specificity ≥ 80.00%. Conclusions: AI Avenue recognizes cases of non-disease very well, while optometrists tend to misdiagnose the grade of disease. Key words: diabetes mellitus, diabetic retinopathy, diabetic macular oedema, automated detection of diabetic retinopathy.