Abstract
Glaucoma\'s irreversibility, lacking of glaucoma specialists and patient unawareness demand for an economic and effective glaucoma diagnosis system for screening. In this study we explore feature selection (FS) technologies to identify the most essential parameters for automatic glaucoma diagnosis. Methods: We compose feature space from heterogeneous data sources, i.e., retinal image and eye screening data. A feature selection framework is proposed by exploring multiple feature ranking schemes and a wide range of supervised learners. The optimal feature set is derived automatically from the performance curve smoothed by measurement score regression. Results: Under the proposed framework, the optimal feature set obtained using mRMR (minimum Redundancy Maximum Relevance) scheme contains only 1/4 of the original features. The classifiers trained upon the optimal feature set are more efficient with better performance in terms of Accuracy and F-score. A detailed investigation on the features in the optimal set indicates that they can be the essential parameters for glaucoma mass screening and image segmentation. Conclusions: An intelligent Computer-aid-diagnosis (CAD) model is constructed for automatic disease diagnosis. The effectiveness of the model is demonstrated in our glaucoma study based on heterogeneous data sets. The effort not only improves the discriminative power, but also facilitates a better understanding of CAD process and may ease the data collection in glaucoma mass screening.
Highlights
MethodsWe compose feature space from heterogeneous data sources, i.e., retinal image and eye screening data
Glaucoma is a chronic and irreversible neurodegenerative eye condition in which the optic nerve fibers and astrocytes are progressively damaged [1,2]
One can use I-S-N-T values to check the compliance of ISNT rule: the normal optic disc usually demonstrates a configuration in which the inferior neuroretinal rim is the widest portion of the rim, followed by the superior rim, and the nasal rim, with the temporal rim being the narrowest portion
Summary
The presented work (as illustrated in Figure 1) is based on two heterogeneous data sets, including screening data from SiMES (Singapore Malay Eye Study) [4] and image data from ORIGA [25] database. For each classifier trained by different feature sets and learning methods, we conducted 10-fold cross validation to measure their Accuracy and F-score; followed by applying regression method to smooth the performance curve; the optimal feature set is detected via first derivative test. We employ 4th degree polynomial curve fitting for regression followed by first derivative and second derivative test to obtain their turning points, which are the optimal feature set sizes for the classifier. 4. In Table 4, mRMR-MIQ outperforms other feature ranking methods in 4 outof 8 classifier, i.e., svm-linear, LR, LDA and KNN. With a compact feature set containing only about 1/4 of the original features, a simpler and faster machine learning method is able to achieve better performance
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
