Researchers provide roadmap for AI in medical imaging

Gaps in knowledge must be filled for artificial intelligence to reach its full potential in radiology.

While artificial intelligence could transform clinical imaging practice over the next decade, research is still in its early stages and knowledge gaps must be filled if AI is to reach its full potential in radiology.

That’s the contention of researchers who have developed a roadmap to help identify and prioritize research needs for academic research laboratories, funding agencies, professional societies, as well as industry.

The roadmap, published on Tuesday in the journal Radiology, is based on an August 2018 workshop held at the National Institutes of Health that addressed the future of AI in medical imaging.

“The scientific challenges and opportunities of AI in medical imaging are profound, but quite different from those facing AI generally,” said lead author, Curtis Langlotz, MD, professor of radiology and biomedical informatics, director of the Center for Artificial Intelligence in Medicine and Imaging, and associate chair for information systems in the Department of Radiology at Stanford University.

“Our goal was to provide a blueprint for professional societies, funding agencies, research labs and everyone else working in the field to accelerate research toward AI innovations that benefit patients,” added Langlotz.

Also See: Most studies evaluating AI in radiology didn’t validate the results

According to the roadmap, key research priorities include:
  • New image reconstruction methods that efficiently produce images suitable for human interpretation from source data.
  • Automated image labeling and annotation methods, including information extraction from the imaging report, electronic phenotyping and prospective structured image reporting.
  • New machine learning methods for clinical imaging data, such as tailored, pre-trained model architectures and distributed machine learning methods.
  • Machine learning methods that can explain the advice they provide to human users (so-called explainable AI).
  • Validated methods for image de-identification and data sharing to facilitate wide availability of clinical imaging data sets.
“Imaging research laboratories are rapidly creating machine learning systems that achieve expert human performance using open-source methods and tools,” states the roadmap. “These artificial intelligence systems are being developed to improve medical image reconstruction, noise reduction, quality assurance, triage, segmentation, computer-aided detection, computer-aided classification, and radiogenomics.”

However, the researchers emphasize that standards bodies, professional societies, governmental agencies and industry must collaborate to accelerate advances in foundational AI research for medical imaging.

More for you

Loading data for hdm_tax_topic #better-outcomes...