Self-supervised microscopic image super-resolution processing method and system

A system for self-supervised microscopy image super-resolution processing, configured to process raw fluorescence image data of a biological sample acquired by an optical imaging system;

• comprising:
• a computer;
• wherein the computer is configured to:
• obtain, depending on a type of the optical imaging system, a training set by performing image preprocessing on the raw fluorescence image data using a computer;
• train, using the training set in the computer, a neural network for performing denoising and super-resolution processing on an image, wherein the neural network comprises a denoising component and a deconvolution component, wherein the training set comprises input image data and ground truth input image data, pixel blocks at a same position are randomly extracted from the input image data and the ground truth input image data and subjected to random rotation and flipping to serve as a training input image and a training target image respectively;
• the training input image is first processed by the denoising component to generate a denoised image, and then processed by the deconvolution component to generate a deconvolved image;
• determine a denoising loss using the denoised image and the training target image, and determine a deconvolution loss based on the deconvolved image, after undergoing degradation processing depending on the type of the optical imaging system, and the training target image;
• wherein during the training process of the neural network, a loss function comprises the denoising loss and the deconvolution loss, or only comprises the deconvolution loss;
• and
• process the raw fluorescence image data using a trained neural network or process additional fluorescence image data of a same biological sample acquired by the optical imaging system using a trained neural network.