UB ScholarWorks

Computer Vision Based Early Intraocular Pressure Assessment From Frontal Eye Images

Show simple item record

dc.contributor.author Al Oudat, Mohammad
dc.date.accessioned 2019-01-08T19:13:09Z
dc.date.available 2019-01-08T19:13:09Z
dc.date.issued 2019-01-04
dc.identifier.citation M.A. Al-Oudat, "Computer Vision Based Early Intraocular Pressure Assessment From Frontal Eye Images", Ph.D. dissertation, Dept. of Computer Science and Engineering, Univ. of Bridgeport, Bridgeport, CT, 2018. en_US
dc.identifier.uri https://scholarworks.bridgeport.edu/xmlui/handle/123456789/3969
dc.description.abstract Intraocular Pressure (IOP) in general, refers to the pressure in the eyes. Gradual increase of IOP and high IOP are conditions or symptoms that may lead to certain diseases such as glaucoma, and therefore, must be closely monitored. While the pressure in the eye increases, different parts of the eye may become affected until the eye parts are damaged. An effective way to prevent rise in eye pressure is by early detection. Exiting IOP monitoring tools include eye tests at clinical facilities and computer-aided techniques from fundus and optic nerves images. In this work, a new computer vision-based smart healthcare framework is presented to evaluate the intraocular pressure risk from frontal eye images early-on. The framework determines the status of IOP by analyzing frontal eye images using image processing and machine learning techniques. A database of images from the Princess Basma Hospital was used in this work. The database contains 400 eye images; 200 images with normal IOP and 200 high eye pressure case images. This study proposes novel features for IOP determination from two experiments. The first experiment extracts the sclera using circular hough transform, after which four features are extracted from the whole sclera. These features are mean redness level, red area percentage, contour area and contour height. The pupil/iris diameter ratio feature is also extracted from the frontal eye image after a series of pre-processing techniques. The second experiment extracts the sclera and iris segment using a fully conventional neural network technique, after which six features are extracted from only part of the segmented sclera and iris. The features include mean redness level, red area percentage, contour area, contour distance and contour angle along with the pupil/iris diameter ratio. Once the features are extracted, classification techniques are applied in order to train and test the images and features to obtain the status of the patients in terms of eye pressure. For the first experiment, neural network and support vector machine algorithms were adopted in order to detect the status of intraocular pressure. The second experiment adopted support vector machine and decision tree algorithms to detect the status of intraocular pressure. For both experiments, the framework detects the status of IOP (normal or high IOP) with high accuracies. This computer vison-based approach produces evidence of the relationship between the extracted frontal eye image features and IOP, which has not been previously investigated through automated image processing and machine learning techniques from frontal eye images. en_US
dc.language.iso en_US en_US
dc.subject Computer vision en_US
dc.subject Eye segmentation en_US
dc.subject Fully convolution network en_US
dc.subject Intraocular pressure en_US
dc.subject Pupil/iris ratio en_US
dc.subject Sclera en_US
dc.title Computer Vision Based Early Intraocular Pressure Assessment From Frontal Eye Images en_US
dc.type Thesis en_US
dc.institute.department School of Engineering en_US
dc.institute.name University of Bridgeport en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search ScholarWorks


Advanced Search

Browse

My Account