Active learning for an efficient training strategy of computer-aided diagnosis systems: application to diabetic retinopathy screening
- PMID: 20879450
- DOI: 10.1007/978-3-642-15711-0_75
Active learning for an efficient training strategy of computer-aided diagnosis systems: application to diabetic retinopathy screening
Abstract
The performance of computer-aided diagnosis (CAD) systems can be highly influenced by the training strategy. CAD systems are traditionally trained using available labeled data, extracted from a specific data distribution or from public databases. Due to the wide variability of medical data, these databases might not be representative enough when the CAD system is applied to data extracted from a different clinical setting, diminishing the performance or requiring more labeled samples in order to get better data generalization. In this work, we propose the incorporation of an active learning approach in the training phase of CAD systems for reducing the number of required training samples while maximizing the system performance. The benefit of this approach has been evaluated using a specific CAD system for Diabetic Retinopathy screening. The results show that (1) using a training set obtained from a different data source results in a considerable reduction of the CAD performance; and (2) using active learning the selected training set can be reduced from 1000 to 200 samples while maintaining an area under the Receiver Operating Characteristic curve of 0.856.
Similar articles
-
A multiple-instance learning framework for diabetic retinopathy screening.Med Image Anal. 2012 Aug;16(6):1228-40. doi: 10.1016/j.media.2012.06.003. Epub 2012 Jul 6. Med Image Anal. 2012. PMID: 22850462
-
Vessel network detection using contour evolution and color components.Annu Int Conf IEEE Eng Med Biol Soc. 2010;2010:3129-32. doi: 10.1109/IEMBS.2010.5626090. Annu Int Conf IEEE Eng Med Biol Soc. 2010. PMID: 21095748
-
Algorithms for digital image processing in diabetic retinopathy.Comput Med Imaging Graph. 2009 Dec;33(8):608-22. doi: 10.1016/j.compmedimag.2009.06.003. Epub 2009 Jul 18. Comput Med Imaging Graph. 2009. PMID: 19616920 Review.
-
Using a patient image archive to diagnose retinopathy.Annu Int Conf IEEE Eng Med Biol Soc. 2008;2008:5441-4. doi: 10.1109/IEMBS.2008.4650445. Annu Int Conf IEEE Eng Med Biol Soc. 2008. PMID: 19163948
-
CAD in questions/answers Review of the literature.Eur J Radiol. 2009 Jan;69(1):24-33. doi: 10.1016/j.ejrad.2008.07.042. Epub 2008 Oct 31. Eur J Radiol. 2009. PMID: 18977103 Review.
Cited by
-
MIRank-KNN: multiple-instance retrieval of clinically relevant diabetic retinopathy images.J Med Imaging (Bellingham). 2017 Jul;4(3):034003. doi: 10.1117/1.JMI.4.3.034003. Epub 2017 Sep 1. J Med Imaging (Bellingham). 2017. PMID: 28894762 Free PMC article.
-
Automated analysis of diabetic retinopathy images: principles, recent developments, and emerging trends.Curr Diab Rep. 2013 Aug;13(4):453-9. doi: 10.1007/s11892-013-0393-9. Curr Diab Rep. 2013. PMID: 23686810 Review.
-
Active learning for clinical text classification: is it better than random sampling?J Am Med Inform Assoc. 2012 Sep-Oct;19(5):809-16. doi: 10.1136/amiajnl-2011-000648. Epub 2012 Jun 15. J Am Med Inform Assoc. 2012. PMID: 22707743 Free PMC article.
MeSH terms
LinkOut - more resources
Medical
Miscellaneous