Sharing AI Lessons Learned

This post was contributed by Alexander M. Norbash, MD, FACR, member of the Radiology Leadership Institute® (RLI) Board and chair of the RLI Summit.

Norbash_AWill artificial intelligence (AI) revolutionize radiology and medicine? Are radiology and AI slated to be future frenemies? So many questions — so few answers at this point.

This is why we may look for clarity from the lessons learned from our pathologist colleagues.

David N. Louis, MD, pathologist-in-chief at Massachusetts General Hospital, Boston, will share how that specialty has approached the opportunities and challenges of AI during the Radiology Leadership Institute® (RLI) Summit.

He will talk about areas in which pathology has connected with AI. These include image-based analysis in addition to understanding how bio-informatics derived from large data sets and laboratory science can benefit patients and society. Radiology and pathology are close siblings; there’s much for us to learn from pathology.

  • How are pathologists dealing with the technological evolution?
  • What have they learned?
  • What recommendations can we take away?

Besides this discussion, Anne Le Grand, general manager of Watson Health Imaging, will explore the potential of cognitive imaging to aid radiologists in delivering high-quality, patient-centered care in “Watson Health: Empowering Heroes, Transforming Health.” Additionally, the Radiology Leadership Institute offers a new podcast from Mark Michalski, MD, on “Artificial Intelligence and the Future of Radiology.”

Last week, Dr. Geraldine McGinty wrote, “Radiology must shape future health care or be shaped by someone else’s vision of 21st century medicine” in talking about the American College of Radiology (ACR) Data Science Institute (DSI).

We should embrace AI and incorporate it into our evolution, rather than seeing it as a competitor; to that end, we need to more fully understand it.

  • What future do you envision with radiology and AI?
  • Will AI increase our efficiency and help radiologists be more valuable?

 Please share your thoughts in the comments section below and join the discussion on Engage (login required).

 

Being “Patient,” Even With AI

This post was contributed by Geraldine B. McGinty, MD, MBA, FACR, vice chair of the American College of Radiology (ACR) Board of Chancellors and chair of the ACR Data Science Institute (DSI) Advisory Group.

Dr.McGinty(updated)_andcroppedArtificial intelligence (AI) and machine learning have the potential revolutionize medicine including radiology. But, the right tools must be developed and implemented to help us improve patient care. The American College of Radiology launched the ACR Data Science Institute (DSI) to do just that.

A key component of the DSI’s success is the creation of the DSI Advisory Group. We are bringing the brightest imaging informatics minds together with industry/technology stakeholders and patient advocates. It’s critical that we include the perspective of our patients, and we are fortunate to have engaged the patient advocate members of the Commission on Patient- and Family-Centered Care Informatics Committee.

The DSI Advisory group also includes a broad spectrum of ACR members — including those from private practice, research and education. Members in training, early-career professionals and those in small and rural practices, radiation oncology and physics are involved. We have members from the Economics, Quality and Safety and Women and Diversity commissions.

Together, we will move the ACR DSI mission to “advance and validate artificial intelligence in medical imaging for the benefit of our patients and profession” forward.

Each Advisory Group member was asked to share an initial perspective on the opportunities and challenges around AI. These include:

  • How do we value augmented physician effort in fee for service payment?
  • Who can use and own patient data?
  • How do we communicate our view for the future?
  • How do we use AI in our practices?

These and other questions will be answered in the coming months — and years. Radiology must shape future health care or be shaped by someone else’s vision of 21st century medicine.

What do you see as opportunities with AI in medical practice?

What are our challenges with AI in medical practice?

Please share your thoughts in the comments section below and join the discussion on Engage (login required).

Evidence Piling Up for Virtual Colonoscopy

Dr. YEEThis post was contributed by Judy Yee, MD, chair of the American College of Radiology Colon Cancer Committee.

A new study in Radiology shows insurance coverage for virtual colonoscopy (CT colonography – CTC) boosts colorectal cancer (CRC) screening nearly 50 percent. This echoes studies in the U.S. and Europe that show improved access to CTC leads to more screening.

This is vital. Colorectal cancer is the second leading cancer killer in men and women. It is most often diagnosed in those ages 65–74. Yet, at least a third of those 50 and older who should be screened choose not to be tested.

Better screening is what the US Preventive Services Task Force (USPSTF) wanted when it assigned an “A” grade to virtual colonoscopy and other CRC screening exams.

With this “A” grade, federal law requires private insurers to cover these exams with no copay. Most already do. Yet, Medicare does not.

The ACR is working with multiple medical societies, patient groups, industry and minority health care advocates to change this. It’s time:

  • Studies in NEJM and elsewhere show it is as accurate as regular colonoscopy in those at average CRC risk. This is most people.
  • It is less expensive than colonoscopy
  • A study of Medicare-aged patients shows no short- or long-term complications from CTC

To make this American Cancer Society recommended screening exam more available, the effort needs data. We also need your help to let patients know that CTC is an insured CRC screening option that they should consider.

Please share your thoughts in the comments section below and join the discussion on Engage (login required).

*This link goes to an industry-maintained database. ACR does not maintain the site.