JBJS - 2026-05-06 - Journal Article; Review
MRI-Based Synthetic CT Shows Promise as a Radiation-Free Alternative to Conventional CT in Orthopaedics.
Stewart H, Watkins A, Ahlawat S, Fayad LM, Skaggs DL, Sponseller PD
Topics
Key Takeaway
MRI-based synthetic CT techniques (ZTE, UTE, 3D-GRE, and deep learning-based sCT) demonstrate high agreement with conventional CT across multiple anatomical sites, with deep learning models showing the strongest improvement trajectory despite limited training datasets.
Summary Depth
Choose how much analysis to show on this article page.
Summary
This narrative/systematic review evaluates MRI-based radiation-free alternatives to CT—specifically ZTE, UTE, 3D-GRE, and deep learning synthetic CT—for orthopaedic bone imaging. Early studies across multiple anatomical sites show high agreement with conventional CT for all modalities, with ZTE offering superior resolution and silent acquisition, and 3D-GRE offering the lowest implementation barrier. Deep learning sCT is identified as the highest-potential modality but is currently constrained by training data volume and anatomical specificity.
Key Limitation
The review lacks quantitative synthesis of agreement data, making it impossible to determine whether any single MRI-based modality meets the threshold for clinical equivalence to CT in high-stakes applications such as oncologic resection planning or complex fracture characterization.
Original Abstract
➢ Computed tomography (CT) remains the gold standard for bone imaging, but radiation risks, especially in children, are driving interest in alternatives. ➢ Magnetic resonance imaging (MRI)-based techniques are emerging as a radiation-free alternative to CT, using sequences such as zero echo time, ultrashort echo time, and 3-dimensional (3D) gradient recalled echo, along with deep learning-based synthetic CT. ➢ Zero echo time MRI stands out for its high-resolution and silent imaging, whereas 3D gradient recalled echo offers widespread availability and minimal requirements for implementation. ➢ Early studies have shown high agreement of all modalities with CT across multiple anatomical sites, supporting broader clinical use, especially in pediatrics, surgical planning, and cost-reduction efforts. ➢ Deep learning-based synthetic CT demonstrates strong potential given its ability to improve over time and to generate highly accurate CT-like images, although current applications are limited by existing training data.