AUTOMATIC OPTIC DISK DETECTION FROM LOW CONTRAST RETINAL IMAGES OF ROP INFANT USING GVF SNAKE

Reliable and efficient optic disk localization and segmentation are important tasks in automated retinal screening. General-purpose edge detection algorithms often fail to segment the optic disk (OD) due to fuzzy boundaries, inconsistent image contrast or missing edge features, especially in infants’ retinal images where the image acquisition process has to be very quick and in low light conditions. This paper presents an algorithm for segmentation of optic disk boundary in low-contrast images. The optic disk localization is achieved using segmentation by a deformable contour model (or Snake) with gradient vector flow (GVF) as an external force. The first Snake is placed at a location very close to the center of the optic disk approximated by a PCA-based model. The algorithm is evaluated using 50 retinal images from infants with retinopathy of prematurity (ROP) condition. The results from the GVF method were compared with conventional optic disk detection using a 2D Circular Hough Transform and later verified with hand-drawn ground truth. The result is quite successful with the accuracy of 85.34%.

[1]  D. T. Morris,et al.  Automated Extraction Of The Optic Nerve Head Rim. , 1993 .

[2]  C. Sinthanayothin,et al.  Automated localisation of the optic disc, fovea, and retinal blood vessels from digital colour fundus images , 1999, The British journal of ophthalmology.

[3]  M J Cox,et al.  Inter‐ and intra‐image variability in computer‐assisted optic nerve head assessment , 1991, Ophthalmic & physiological optics : the journal of the British College of Ophthalmic Opticians.

[4]  Shaogang Gong,et al.  Dynamic Vision - From Images to Face Recognition , 2000 .

[5]  Junaed Sattar.  Snakes , Shapes and Gradient Vector Flow , 2022 .

[6]  K.S. Park,et al.  Three dimensional reconstruction of conventional stereo optic disc image , 2004, The 26th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[7]  W. Eric L. Grimson,et al.  Utilizing Segmented MRI Data in Image-Guided Surgery , 1997, Int. J. Pattern Recognit. Artif. Intell..

[8]  Huiqi Li,et al.  Automatic location of optic disk in retinal images , 2001, Proceedings 2001 International Conference on Image Processing (Cat. No.01CH37205).

[9]  Langis Gagnon,et al.  Fast and robust optic disc detection using pyramidal decomposition and Hausdorff-based template matching , 2001, IEEE Transactions on Medical Imaging.

[10]  Hideki Kuga,et al.  A computer method of understanding ocular fundus images , 1982, Pattern Recognit..

[11]  Enrico Grisan,et al.  Detection of optic disc in retinal images by means of a geometrical model of vessel structure , 2004, IEEE Transactions on Medical Imaging.

[12]  B M Dawant,et al.  Brain segmentation and white matter lesion detection in MR images. , 1994, Critical reviews in biomedical engineering.

[13]  Lena Costaridou,et al.  Medical Image Analysis Methods , 2005 .

[14]  Demetri Terzopoulos,et al.  On Matching Deformable Models to Images , 1987, Topical Meeting on Machine Vision.

[15]  Jerry L. Prince,et al.  Gradient vector flow: a new external force for snakes , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[16]  C. Sinthanayothin,et al.  images retinal blood vessels from digital colour fundus Automated localisation of the optic disc , fovea , and , 1999 .