Updated
Updated · MIT News · Sep 16
MIT Researchers Build 5-Minute AI Tool for Sub-Millimeter Surgical X-Ray Matching
Updated
Updated · MIT News · Sep 16

MIT Researchers Build 5-Minute AI Tool for Sub-Millimeter Surgical X-Ray Matching

1 articles · Updated · MIT News · Sep 16

Summary

  • Nature published MIT’s xvr system, which adapts to each patient in about five minutes and then aligns intraoperative X-rays with preoperative CT or MRI scans in seconds.
  • Sub-millimeter registration aims to make minimally invasive procedures safer and faster by showing clinicians exactly where catheters, endoscopes and other tools sit inside the body.
  • Xvr generates roughly 1,000 synthetic X-ray images per second from a patient’s 3D scan, using physics-based simulation to train a patient-specific model without relying on broad one-size-fits-all predictions.
  • More than 2,000 whole-body scans were used to pretrain a foundation model, and tests on data from five hospitals showed xvr beat existing AI methods by an order of magnitude across varied patients and procedures.
  • The team is now working on real-time deployment, moving-body scenarios and partnerships with surgical robotics and clinical groups, with potential use in emergency interventions such as stroke care.

Insights

How does a new AI system use synthetic physics to turn flat surgical X-rays into sub-millimeter 3D maps in seconds?
Could training AI on simulated data rather than real patients be the secret to eliminating errors in minimally invasive surgeries?