• Politics
  • Diversity, equity and inclusion
  • Financial Decision Making
  • Telehealth
  • Patient Experience
  • Leadership
  • Point of Care Tools
  • Product Solutions
  • Management
  • Technology
  • Healthcare Transformation
  • Data + Technology
  • Safer Hospitals
  • Business
  • Providers in Practice
  • Mergers and Acquisitions
  • AI & Data Analytics
  • Cybersecurity
  • Interoperability & EHRs
  • Medical Devices
  • Pop Health Tech
  • Precision Medicine
  • Virtual Care
  • Health equity

AI Tool Identifies, Locates Aneurysms from Brain Scans

Article

The model increased the clinicians’ accuracy by 0.038.

brain scan

Photo/Thumb have been modified. U.S. Army Photo.

An artificial intelligence (AI) tool can highlight areas of a brain scan that are likely to contain an aneurysm, according to the findings of a new study published in JAMA Network Open.

The tool, built around an algorithm called HeadXNet, improved clinicians’ ability to correctly identify aneurysms at a rate equivalent to finding six more aneurysms in 100 scans of aneurysms. The clinicians’ mean sensitivity increased by 0.059, the mean accuracy increased by 0.038 and the mean interrater agreement increased by 0.060.

Using the AI tool also improved consensus among the interpreting clinicians.

“There’s been a lot of concern about how machine learning will actually work within the medical field,” said Allison Park, co-lead author of the paper and a graduate student in statistics at Stanford University. “This research is an example of how humans stay involved in the diagnostic process, aided by an (AI) tool.”

The researchers developed and applied HeadXNet to augment clinicians’ intracranial aneurysm diagnostic performance.

Kristen Yeom, M.D., associate professor of radiology and co-senior author, and Andrew Ng, Ph.D., adjunct professor of computer science and co-senior author, developed a 3D convolutional neural network architecture using a training set of 611 head computed tomographic angiography examinations to generate aneurysm segmentations. Researchers provided a test set of 115 examinations to clinicians.

Eight physicians diagnosed for aneurysm on the test set with and without the model.

Using HeadXNet, clinicians correctly identified more aneurysm, reduced the “miss” rate and were more likely to agree with their colleagues. Not only did the algorithm say whether the scan contained an aneurysm, but it also helped pinpoint the exact locations of the aneurysms.

“Search for an aneurysm is one of the most labor-intensive and critical tasks radiologists can undertake,” Yeom said. “Given inherent challenges of complex neurovascular anatomy and potential fatal outcome of a missed aneurysm, it prompted me to apply advances in computer science and vision to neuroimaging.”

Get the best insights in digital health directly to your inbox.

Related

AI Model Shows Promise in Predicting Lung Cancer, Google Study Finds

AI Directs Catheters to Target to Treat A-Fib

FDA Clears Aidoc's AI Solution for Triage of Cervical Spine Fractures

Recent Videos
Image: Ron Southwick, Chief Healthcare Executive
George Van Antwerp, MBA
Edmondo Robinson, MD
Craig Newman
Related Content
© 2024 MJH Life Sciences

All rights reserved.