Media.Player

A Novel Comprehensive 3D Analysis with Spectral Domain OCT

Detailed Technology Description

None

*Abstract

BackgroundRetinal images take an important role in diagnosis of ocular diseases, such as glaucoma, diabetic retinopathy (DR) and age-related macular degeneration (AMD). They are major causes of blindness in the US and throughout the world. As majority of patients do not experience any symptoms until the disease reaches an advanced stage, early detection and treatment is of utmost importance. The legacy Optical Coherence Tomography (OCT) methods show limitations in detecting subtle structural changes, which is important in early detection of ocular diseases and the ability to track longitudinally the disease dynamic. Technology3D SD-OCT retinal images are composed of a series of cross-sectional scans from top to bottom of the scanning region on retina. Each B-scan consists of certain number of high-resolution one-dimensional scans in the z direction. The developed 3D analysis technique uses self-size-adjusting super pixel segmentation and machine classifier to quantitatively summarize the full 3D dataset, assess subtle structural damages, and automatically identify the ocular diseases. There are 3 steps in the present invention. A detailed description of the procedures is given as follows:1. Pre-process the 3D SD-OCT volume data2. Self-size-adjusting super pixel segmentation3. Automatically identify ocular diseases and disease progression using super pixel features and machine classifier.The proposed technique is robust and scalable for various ocular diseases. As super pixel boundary definition can be defined by taking different sets of features, optimized targeting of a specific disease can be easily accomplished.Application* Diagnosis of ocular diseases. * Clinical glaucoma management.Advantages1. Fast scanning speed allows three-dimensional (3D) volume scanning.2. Fast scanning also provides crucial information for structural assessment of various eye diseases.3. Self-size-adjusting super pixel segmentation.4. Improved ability of detecting subtle structural changes, thus improve ocular disease discrimination and its progression detection. 5. Robust and scalable for various ocular diseases.Stage of Development* Software developed and preliminary testing complete.Copyright

*Principal Investigator

Name: Hiroshi Ishikawa, Associate Professor, Director

Department: Med-Ophthalmology

Name: Joel Schuman, Professor & Chairman of Ophthalmology

Department: Med-Ophthalmology

Name: Chaim-Gadi Wollstein

Department: Med-Ophthalmology

Name: Juan Xu, Research scholar

Department: Med-Ophthalmology

Country/Region

USA