Signal to-noise enhancement in imaging applications using a time-series of images

A method of reducing noise in an image stack, which includes a time-series of images, comprising:

• capturing, using imaging hardware, multiple images of a subject using a single image protocol, wherein the multiple images are captured with an amount of time between images;
• performing a localized analysis using a sub-region of the multiple images in order to remove components with lower amplitudes that are within the sub-region, wherein a sub-region includes two or more pixels in each of the multiple images;
• moving the sub-region around the multiple images and performing, using a computer, the localized analysis independently on each sub-region until all pixels in the multiple images have been analyzed to obtain multiple filtered images with components removed;
• and
• using the computer, combining the multiple filtered images to obtain a single final image from the image stack.

A method of reducing noise in an image stack, which includes a time series of images, comprising:

• capturing a time series of images using an imaging modality;
• using a computer, performing a localized component analysis on sub-regions projected across each image in the time series, each sub-region including a plurality of pixels;
• canceling at least some of the components on a localized basis to obtain a series of images with reduced components;
• combining the series of images with reduced components into a single image by analyzing each corresponding pixel across the images and determining a pixel value for each pixel based on the analysis.