Medical image super-resolution reconstruction method and system

Abstract

The invention discloses a medical image super-resolution reconstruction method and system. A distillation progressive refinement module is constructed, and hierarchical structure features are gradually extracted, and the most useful feature information can be gradually screened out through the cooperation of contrast-aware attention modules. This can not only extract deep features, but also retain high-frequency parts of the image, such as edges, structures, etc. Finally, through the combined up-sampling module, the low-resolution medical image can be reconstructed into a super-resolution image with clear edges and detailed information. Improve the efficiency of medical image super-resolution (the time to reconstruct a single image is about 38 milliseconds). Experiments have proved that the medical image super-resolution method based on multi-feature distillation of the present invention can improve the quality and efficiency of medical super-resolution images. The objective indicators and reconstruction time of the images are better than the comparison methods, and the parameters of the model are relatively large. It is about 6 times less than the comparison model.

Description

technical field [0001] The invention relates to image processing technology, in particular to a medical image super-resolution reconstruction method and system. Background technique [0002] With the development of computer technology and modern medical technology, artificial intelligence technology is widely used in the medical field. Confocal Laser Scanning Microscope-Generated Dermatology Images Classified by a Convolutional Neural Network-Based Diagnostic Aid Model Almost as Accurately as a Dermatologist's Classification [1] . On the basis of CNN, American Telecom has developed telemedicine. Doctors can communicate with patients in need through online video communication, and provide patients with diagnosis and treatment options; Patients with major diseases such as critical diseases provide expert medical advice and diagnosis from top medical institutions in the United States; under the new crown pneumonia epidemic, the United States has fully opened up patients to re...

AI Technical Summary

Problems solved by technology

these methods [6][7][8][9][10] Using various methods such as residual completion, recursive learning, and gradient learning to learn more information about the input image, although the effect of image super-resolution reconstruction is improved, but computational efficiency is sacrificed, quality and efficiency cannot be considered, and it is not suitable for direct For super-resolution reconstruction of medical images

Images