Sonic DL is a software feature for GE HealthCare MR systems that uses deep learning algorithms to reconstruct MR images from highly under-sampled data, enabling faster image acquisition and maintaining diagnostic quality. It supports multiple MR field strengths and is used across various anatomies, including cardiac, brain, spine, musculoskeletal, abdomen, pelvis, and breast imaging, helping clinicians obtain high-quality MR images more quickly.
Sonic DL is a Deep Learning based reconstruction technique for use on GE HealthCare 1.5T, 3.0T, and 7.0T MR systems to reconstruct MR images from highly under-sampled data, enabling highly accelerated acquisitions for patients of all ages across all anatomies.
Sonic DL employs a deep learning convolutional neural network integrated into GE HealthCare MR system software, reconstructing images from under-sampled 2D cardiac cine acquisitions and 3D Cartesian acquisitions. It uses fast spin echo and gradient echo sequences and is compatible with MR field strengths of 1.5T, 3.0T, and 7.0T. It achieves image acceleration factors up to 12, outperforming conventional parallel imaging and compressed sensing methods.
Extensive nonclinical and clinical testing demonstrated Sonic DL 3D can accelerate MR acquisitions up to 12-fold, maintaining diagnostic image quality across varied anatomies and field strengths. Nonclinical tests showed superior image quality to conventional methods while clinical studies confirmed diagnostic image quality and pathology retention. Quantitative analysis indicated minimal differences to standard reconstructions, and stability testing showed low risk of hallucination or artifacts. Recommended acceleration factors vary by anatomy and field strength, with some tradeoff in low contrast detectability at very high acceleration.
No predicate devices specified
Submission
11/27/2024
FDA Approval
6/5/2025
Join hundreds of your peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.
We respect your privacy. Unsubscribe at any time.