Learning Apparent Diffusion Coefficient Maps from Undersampled Radial k-Space Diffusion-Weighted MRI in Mice using a Deep CNN-Transformer Model in Conjunction with a Monoexponential Model: Paper and Code

Get our free extension to see links to code for papers anywhere online!Free add-on: code for papers everywhere!Free add-on: See code for papers anywhere!

Add to Chrome

Add to Firefox

Add to Edge

Learning Apparent Diffusion Coefficient Maps from Undersampled Radial k-Space Diffusion-Weighted MRI in Mice using a Deep CNN-Transformer Model in Conjunction with a Monoexponential Model

Jul 06, 2022
Yuemeng Li, Hee Kwon Song, Miguel Romanello Joaquim, Stephen Pickup, Rong Zhou, Yong Fan

Figure 1 for Learning Apparent Diffusion Coefficient Maps from Undersampled Radial k-Space Diffusion-Weighted MRI in Mice using a Deep CNN-Transformer Model in Conjunction with a Monoexponential Model

Figure 2 for Learning Apparent Diffusion Coefficient Maps from Undersampled Radial k-Space Diffusion-Weighted MRI in Mice using a Deep CNN-Transformer Model in Conjunction with a Monoexponential Model

Figure 3 for Learning Apparent Diffusion Coefficient Maps from Undersampled Radial k-Space Diffusion-Weighted MRI in Mice using a Deep CNN-Transformer Model in Conjunction with a Monoexponential Model

Figure 4 for Learning Apparent Diffusion Coefficient Maps from Undersampled Radial k-Space Diffusion-Weighted MRI in Mice using a Deep CNN-Transformer Model in Conjunction with a Monoexponential Model

Share this with someone who'll enjoy it:

Purpose: To accelerate radially sampled diffusion weighted spin-echo (Rad-DW-SE) acquisition method for generating high quality of apparent diffusion coefficient (ADC) maps. Methods: A deep learning method was developed to generate accurate ADC map reconstruction from undersampled DWI data acquired with the Rad-DW-SE method. The deep learning method integrates convolutional neural networks (CNNs) with vison transformers to generate high quality ADC maps from undersampled DWI data, regularized by a monoexponential ADC model fitting term. A model was trained on DWI data of 147 mice and evaluated on DWI data of 36 mice, with undersampling rates of 4x and 8x. Results: Ablation studies and experimental results have demonstrated that the proposed deep learning model can generate high quality ADC maps from undersampled DWI data, better than alternative deep learning methods under comparison, with their performance quantified on different levels of images, tumors, kidneys, and muscles. Conclusions: The deep learning method with integrated CNNs and transformers provides an effective means to accurately compute ADC maps from undersampled DWI data acquired with the Rad-DW-SE method.

View paper on

Share this with someone who'll enjoy it: