This week saw a number of new announcements from vendors across a range of AI use cases: improving population health management at the county level and a New York University Langone Health subsidiary focused on leveraging physics to improve AI-powered imaging technology.
Additionally, with official approval from the U.S. Food and Drug Administration, the first imaging software could give clinicians access to a powerful denoising algorithm that improved the quality of magnetic resonance images in a recent neuroimaging study by developer Microstructure Imaging.
FDA Approves RMT for Clinical Practice
Microstructure Imaging, a Brooklyn, New York-based vendor that specializes in MRI software development, announced Wednesday that it has received FDA 501 clearance for a breakthrough in signal-to-noise diffusion and functional MRI software.
While MRI machines are often noisy for patients, the noise generated by the images is a bigger concern for their overall care. MRI uses powerful magnetic fields and radio waves to take detailed images of organs and tissues, but the clarity of the images depends on the signal-to-noise ratio (SNR), a calculation that shows the difference between the area of interest and the background at each pixel.
While more medical insight can be gained from images with higher SNRs, MICSI says it leveraged the addition of a random matrix theory algorithm to deliver a “multiplying increase in SNR.”
The software, called MICSI-RMT, is the first to bring RMT denoising technology to clinical practice, the company said, making it particularly transformative for neuroimaging applications, such as in stroke management, where accurate imaging improves medical decision-making.
The software can improve the quality of MRI images without requiring external training data or high-performance GPUs, unlike traditional AI-enhanced imaging, the company said in a statement.
According to the company’s website, that means the AI-enhanced denoising method is not hallucinogenic.
MICSI-RMT also handles modules that support weighted linear least squares and Bayesian matching techniques to facilitate quantitative analysis of diffusion-weighted images, and notes that DICOM data routing ensures secure integration into medical workflows.
In the company’s blinded evaluation study, neurologists evaluated MRI images processed with MICSI-RMT and standard-of-care tools and said they preferred RMT images for clarity and reduced noise. They also said MICSI-RMT-processed DTI images had sharper images of small structures in the brain’s white matter and improved contrast with neighboring tissue types. They also reported activation maps that were more anatomically consistent with functional MRI images generated with RMT technology.
The researchers found that RMT technology improved diffusion MRI by 4.35 times and fMRI by 1.9 times, providing more accurate parametric maps. There was also a significant improvement – 56.3% – in dMRI enhancement with ADC map accuracy.
District strengthens patient analysis
Innovaccer Inc., a healthcare AI platform company, announced Wednesday that the San Mateo County Health Department will use its population health management platform to improve patient care and data management for approximately 165,000 people and 500,000 seniors across the county.
By compiling data across SMCH systems and integrating it into the electronic health record, healthcare providers can search for patients across multiple EHR systems, review longitudinal patient records, and transfer data back to their primary EHR, the company said.
“This partnership is a great example of teams successfully working together to seamlessly integrate data and improve care delivery while bringing joy to care,” Abhinav Shashank, co-founder and CEO of Innovaccer, said in a statement.
The HIMSS Healthcare AI Forum is scheduled for September 5-6 in Boston. Learn more and register.
Andrea Fox is senior editor of Healthcare IT News.
Email: [email protected]
Healthcare IT News is a publication of HIMSS Media.
