Fetal Cardiac Cine MRI: Deep Learning Reconstruction at 1.5T

Same fetal cardiac acquisition. Look at what changes after deep learning reconstruction.

Fetal cardiac cine is one of the most challenging CMR exams

With a heart roughly the size of a strawberry and a rapid rate of 120 to 160 bpm, fetal cardiac cine imaging presents severe technical constraints. Gating, fetal movement and maternal discomfort make it a difficult exam, ultimately leading to decreased image quality.

Under these circumstances, compromises are necessary: larger voxels preserve SNR, and shorter acquisitions can help mitigate motion artifacts. Even with an optimized protocol, standard acquisitions can result in grainy images with poorly defined myocardial borders.

In this case, SwiftMR® deep learning image reconstruction was combined with Northh Medical’s Smart-Sync Doppler ultrasound gating, which detects and synchronizes the fetal heartbeat with the MR system.

Image Details

Scanner Siemens MAGNETOM Aera 1.5T
Exam Fetal cardiac cine
Gating Northh Medical Smart-Sync Doppler ultrasound gating
Acquired resolution 1.4 × 1.7 × 3.0 mm
SwiftMR reconstructed resolution 0.26 × 0.26 × 3.0 mm

Fetal cardiac cine MRI on a Siemens MAGNETOM Aera 1.5T, conventional reconstruction on the left and SwiftMR reconstruction on the right.

Left: conventional reconstruction. Right: the same acquisition processed with SwiftMR.

Look at the ventricular septum, valve leaflets, and the blood-myocardium boundary. Noise has decreased and the edges are sharper, which is exactly where fetal cardiac reads get uncertain.

Learn More

Explore more SwiftMR clinical images in the AIRS Medical clinical image gallery.

For additional clinical examples or protocol discussions, contact Anthony Rodenberg, BA, R.T.(R)(MR)(ARRT), Director, Clinical Programs.