MIT researchers’ new VoxelMorph machine learning algorithm compares medical images more than 1,000 times faster than current algorithms

June 21, 2018

  • Faster Image Analysis – MIT researchers developed machine learning algorithm, called VoxelMorph, that can compare brain scans and other 3D images more than 1,000 times faster than traditional systems
  • Speed Comparison – Successfully compared 250 test brain scans in two seconds using central processing unit (CPU), and one second using graphics processing unit (GPU), compared to two hours with current algorithms
  • AI Training – Algorithm learned from 7,000 publicly available brain scans, enabling it to align images and estimate optimal alignment parameters
  • Standard-Level Accuracy – VoxelMorph found to be as accurate as commonly used algorithms
  • Other Applications – Can be used to analyze lung images, as well as comparing images before and after surgical operations
Sector: Healthcare/Health Sciences
Organization: Massachusetts Institute of Technology

Original Publication Date: June 18, 2018

What are your thoughts?

AcceleratingBiz® is a trademark of MangoStrategy, LLC


   +1 617-588-3400
Become part of the community!

Receive the latest AcceleratingBiz updates and access member-only content

© 2013-2020 MangoStrategy, LLC   |   Read our Cookie Policy, Privacy Policy and Terms of Service