Image fusion under Multiframe Super-resolution framework

This code implements a method to fuse medical images generated from multiple imaging modalities, such as CT, MRI, and SPECT.

In this work, we have demonstrated that the multiframe super-resolution framework can be used in medical image fusion applications. The proposed method aggregates information from input images generated by imaging modalities (e.g., CT, MRI, and SPECT) to produce a high-resolution image. One advantage of our approach is that it simultaneously perform fusion, enhancement of spatial resolution, and suppression of unwanted featured introduced into the input images. Compared with other methods, specifically Deep Convolutional Neural Network, Principal Component Analysis, and Wavelet), our method demonstrates competitive values of entropy, peak signal-to-noise ratio, and structural similarity.

The corresponding paper for this work has been submitted to EURASIP Journal on Advances in Signal Processing, and is currently under peer review.