• AI
  • Molecular Imaging
  • CT
  • X-Ray
  • Ultrasound
  • MRI
  • Facility Management
  • Mammography

Using Computer-Based Image Retrieval for Radiology Decision Support

Article

CHICAGO - Imagine reading a CT scan of a liver tumor and being able to easily view similar scans, diagnoses, or tumors. It’s a vision of Sandy Napel, PhD, professor of radiology at Stanford School of Medicine, who has been working with his colleagues to develop an image analysis and decision support tool.

CHICAGO - Imagine reading a CT scan of a liver tumor and being able to easily view similar scans, diagnoses, or tumors. It’s a vision of Sandy Napel, PhD, professor of radiology at Stanford School of Medicine, who has been working with his colleagues to develop an image analysis and decision support tool.

“How many times have you been looking at something and saying, ‘When was the last time I saw a liver tumor like this?’” Napel asked the audience at an RSNA session Wednesday.

His pilot system, known as the Electronic Physician Annotation Device (ePAD), characterizes images with features and semantic annotations that can be accessed by the computer for comparison to other images based on a feature vector that scores and reorders the images based on similarity. The database uses consistent radiology terms and features familiar to radiology to then input and organize the images.

For example, a radiologist can query the system based on shape of the lesion or texture, bringing up a host of similar images and information to inform the current read. The information can be used in decision support through computer-based image retrieval.

“This can be made to work and hopefully there will be a time in the near future where you will be able to look at images similar to the patient you are trying to evaluate,” he said, adding that it should be applicable in other disease and imaging scenarios. His group is now looking at how include genomic and other clinical data to images of non-small cell lung cancer to aid in diagnosis and treatment.

Such a computer-based image retrieval system will provide similar images, similar diagnoses, and even evidence-based, relevant data to help guide the radiologist in decision support. Napel added, “This will all be moved into PACS 2.0.”

Recent Videos
Can Diffusion Microstructural Imaging Provide Insights into Long Covid Beyond Conventional MRI?
Assessing the Impact of Radiology Workforce Shortages in Rural Communities
Emerging MRI and PET Research Reveals Link Between Visceral Abdominal Fat and Early Signs of Alzheimer’s Disease
Reimbursement Challenges in Radiology: An Interview with Richard Heller, MD
Nina Kottler, MD, MS
Related Content
© 2024 MJH Life Sciences

All rights reserved.