AI speeds MRI scans

Facebook and NYU’s fastMRI project, led by Larry Zitnick, uses AI in an attempt to make MRI imaging 10 times faster. Neural networks will be trained to fill in missing or degraded parts of scans, turning them from low resolution into high. The goal is to significantly reduce the time patients must lie motionless inside an MRI machine.

Join ApplySci at the 9th Wearable Tech + Digital Health + Neurotech Boston conference on September 24, 2018 at the MIT Media Lab.  Speakers include:  Rudy Tanzi – Mary Lou Jepsen – George ChurchRoz PicardNathan IntratorKeith JohnsonJuan EnriquezJohn MattisonRoozbeh GhaffariPoppy Crum – Phillip Alvelda Marom Bikson – Ed Simcox – Sean Lane

Ultra-low-field, portable MRI

Los Alamos National Laboratory‘s Michelle Espy is developing an ultra-low-field, lightweight MRI system for use on the battlefield and in poor countries.  The device will be simple to transport, set up, and use in non-traditional settings.

Conventional MRI machines use large magnetic fields that align protons in water molecules. Magnetic resonance signals are detected and turned into images. Highly detailed images are created, but the process is complicated and expensive. Espy uses Superconducting Quantum Interference Devices (SQUID) to create quality images with ultra-low-magnetic fields.

The  first generation (battlefield) “b”MRI was built in a large metal housing to shield it from interference.  The team is now surrounding the system with lightweight wire coils in the open environment to compensate the Earth’s magnetic field.  A field compensation system will soon eradicate invading magnetic field signals.


High speed MRI analyzes vocal movement

Aaron Johnson of The Beckman Institute has developed an MRI technique that can view dynamic images of vocal movement at 100 frames per second.  This speed is far more advanced than any other MRI technique.  The method is especially useful in studying how rapidly the tongue moves, along with other muscles in the head and neck, during speech and singing.  The attached video demonstrates the results.

To combine the imaging with audio, the researchers used a noise-canceling fiber-optic microphone to pull out the voice, and aligned the audio track with the imaging.

According to Johnson, the neuromuscular system and larynx change and atrophy with age, contributing to deficits associated with the older voice, such as a weak, strained, or breathy voice.  He wants to understand how these changes occur, and if interventions, such as vocal training, can reverse the effects. This requires seeing how the muscles of the larynx move in real time.

Wearable Tech + Digital Health NYC 2015 – June 30 @ New York Academy of Sciences.  Early registration discount ends Friday, 4/24.