This paper presents a combined approach to automatic extraction of blood vessels in retinal images. The proposed procedure is composed of two phases: a wavelet transform-based preprocessing phase and a NN-based one. Several neural net topologies and training algorithms are considered with the aim of selecting an effective combined method. Human retinal fundus images, derived from the publicly available ophthalmic database DRIVE, are processed to detect retinal vessels. The approach is tested by considering performances in terms of sensitivity and specificity values obtained from vessel classification. The quality of vessel identifications is evaluated on obtained image by computing both sensitivity values and specificity ones and by relating them in ROC curves. A comparison of performances by ROC curve areas for various methods is reported.
Retinal Vessel Extraction by a Combined Neural Network-Wavelet Enhancement Method / Carnimeo, Leonarda; Bevilacqua, Vitoantonio; Cariello, Lucia; Mastronardi, Giuseppe (LECTURE NOTES IN COMPUTER SCIENCE). - In: Emerging intelligent computing technology and applications : with aspects of artificial intelligence : 5th International Conference on Intelligent Computing, ICIC 2009 / [a cura di] De-Shuang Huang; Kang-Hyun Jo; Hong-Hee Lee; Hee-Jun Kang; Vitoantonio Bevilacqua. - STAMPA. - Berlin; Heidelberg : Springer, 2009. - ISBN 978-3-642-04019-1. - pp. 1106-1116 [10.1007/978-3-642-04020-7_118]
Retinal Vessel Extraction by a Combined Neural Network-Wavelet Enhancement Method
Leonarda Carnimeo;Vitoantonio Bevilacqua;Giuseppe Mastronardi
2009-01-01
Abstract
This paper presents a combined approach to automatic extraction of blood vessels in retinal images. The proposed procedure is composed of two phases: a wavelet transform-based preprocessing phase and a NN-based one. Several neural net topologies and training algorithms are considered with the aim of selecting an effective combined method. Human retinal fundus images, derived from the publicly available ophthalmic database DRIVE, are processed to detect retinal vessels. The approach is tested by considering performances in terms of sensitivity and specificity values obtained from vessel classification. The quality of vessel identifications is evaluated on obtained image by computing both sensitivity values and specificity ones and by relating them in ROC curves. A comparison of performances by ROC curve areas for various methods is reported.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
