Abstract: Embodiments of the present invention provide a system and method to automatically screen images from a plurality of fundus cameras for a plurality of retinal diseases. The system comprises stacked autoencoders, amplitude modulation-frequency modulation filters, and artificial neural networks. An embodiment of the present invention uses a multi-class classifier, such as a support vector machine (multi-class SVM) classifier, to determine a plurality of retinal pathologies. An exemplary embodiment of the present invention is validated on a proprietary database of thousands of images (for example over 250,000 retinal images) from multiple cameras, ranging from the table-top fundus cameras to portable hand-held cameras.

Abstract: Systems and methods of obtaining and recording fundus images by minimally trained persons, which includes a camera for obtaining images of a fundus of a subject's eye, in combination with mathematical methods to assign real time image quality classification to the images obtained based upon a set of criteria. The classified images will be further processed if the classified images are of sufficient image quality for clinical interpretation by machine-coded and/or human-based methods. Such systems and methods can thus automatically determine whether the quality of a retinal image is sufficient for computer-based eye disease screening. The system integrates global histogram features, textural features, and vessel density, as well as a local non-reference perceptual sharpness metric. A partial least square (PLS) classifier is trained to distinguish low quality images from normal quality images.

Abstract: Systems and methods of obtaining and recording fundus images by minimally trained persons, which includes a camera for obtaining images of a fundus of a subject's eye, in combination with mathematical methods to assign real time image quality classification to the images obtained based upon a set of criteria. The classified images will be further processed if the classified images are of sufficient image quality for clinical interpretation by machine-coded and/or human-based methods. Such systems and methods can thus automatically determine whether the quality of a retinal image is sufficient for computer-based eye disease screening. The system integrates global histogram features, textural features, and vessel density, as well as a local non-reference perceptual sharpness metric. A partial least square (PLS) classifier is trained to distinguish low quality images from normal quality images.

Abstract: A system and method for assessing a peripheral neuropathy risk wherein the system and method according to one embodiment of the present invention detects abnormality which may be directed to a probability, a stage, a category, or other indicator as to the degree of progression of the disease.