New publication in Magnetic Resonance in Medicine

Symbolic picture for the article. The link opens the image in a large view.

We are happy to announce that our latest publication “CineVN: Variational network reconstruction for rapid functional cardiac cine MRI” (https://onlinelibrary.wiley.com/doi/10.1002/mrm.30260) has been published in Magnetic Resonance in Medicine!

We demonstrate fast inline reconstructions of high-resolution real-time cine MRI with better image quality and more accurate evaluation of cardiac function and myocardial strain compared to state-of-the-art compressed sensing reconstructions. This research was a joint effort of the Computational Imaging Lab with Siemens Healthineers AG, MVZ Diagnostikum Berlin GmbH, and The Ohio State University.

The proposed method is based on a Variational Network architecture adapted for spatiotemporal data and combines it with ideas from conjugate gradient descent.

Architecture of the CineVN consisting of N=15 cascades. Each cascade contains a data consistency (DC) and a refinement (R) term. After every third cascade, we insert a conjugate gradient (CG) block. Trainable parameters are denoted in red.

 

Short-axis slice from a healthy volunteer acquired using the prospective two-shot and real-time protocols, reconstructed using CineVN, Vendor compressed sensing (CS), and CS reconstruction with temporal total variation regularization (TTV-CS). Red arrows denote areas were improved temporal fidelity is particularly visible. Yellow arrows show an area with residual aliasing in the CS reconstruction methods and most notably in the TTV-CS reconstruction.