Abstract: Methods, systems, and computer readable media are provided for reduced-dose angiography using machine learning (e.g., deep learning). Briefly, techniques described herein may use a neural network trained to conserve/preserve angiographic image quality while reducing the angiographic dose of potentially harmful chemical contrast and/or x-ray radiation. As a result, angiographic anatomy may be extracted from an image at reduced angiographic doses using a deep learning neural network. The reduction in chemical contrast and x-ray dose may be achieved based on operations performed before, during, and/or after angiographic imaging.

Abstract: Methods and systems are provided for extracting cardiac frequency angiographic phenomena for an unconstrained vascular object from an angiographic study. In one example, a computer may obtain a series of angiographic image frames obtained at a rate faster than cardiac frequency. Each image frame may comprise a plurality of pixels, and each pixel may have a corresponding intensity. The computer may apply an optical flow technique to the angiographic image frames to generate a plurality of paths corresponding to a displacement of respective pixels from image frame to image frame. The computer may further generate a spatiotemporal reconstruction of cardiac frequency angiographic phenomena based on the plurality of paths and the corresponding intensities associated with respective pixels of the paths, and output for display the spatiotemporal reconstruction of cardiac frequency angiographic phenomena in one or more images.