TIST-Net: style transfer in dynamic contrast enhanced MRI using spatial and temporal information

In this study, our objective is to develop a style transfer method that incorporates spatio-temporal information to either add or remove contrast enhancement from an existing image.
 
We propose a Temporal Image-to-Image Style Transfer Network (TIST-Net), consisting of an auto-encoder combined with convolutional long short-term memory (LSTM) networks. This enables disentanglement of the content and style latent spaces of the time series data, using spatio-temporal information to learn and predict key structures . To generate new images , we use deformable and adaptive convolutions which allow fine grained control over the combination of the content and style latent spaces. We evaluate our method, using popular metrics and a previously proposed contrast weighted structural similarity index measure (CW-SSIM). We also perform a clinical evaluation, where experts are asked to rank images generated by multiple methods. 
 
Our model achieves state-of-the-art performance on three datasets (kidney, prostate and uterus) achieving an SSIM of 0.91±0.03, 0.73±0.04, 0.88±0.04 respectively when performing style transfer between a non-enhanced image and a contrast-enhanced image. Similarly, SSIM results for style transfer from a contrast-enhanced image to a non-enhanced image were 0.89±0.03, 0.82±0.03, 0.87±0.03. In the clinical evaluation, our method was ranked consistently higher than other approaches.
 
TIST-Net can be used to generat...
Source: Physics in Medicine and Biology - Category: Physics Authors: Source Type: research