Skip to content

Editorial: Radiology 2040

Key findings

  • This editorial aims to help radiologists prepare for profound operational and technological developments in the field over the next two decades so they can remain important members of clinical teams
  • Home hospital care, portable imaging, patient self-examination, and teleradiology will continue to expand
  • The integration of artificial intelligence into radiology will continuously evolve, and radiologists must ensure they add value beyond machine-generated interpretations and recommendations
  • The ability to offer therapy will also be essential for strengthening radiologists' clinical impact, so the field needs to cultivate interventional radiology and radiotheranostics extensively

The next two decades are sure to bring "seismic" changes to healthcare, according to James A. Brink, MD, radiologist-in-chief at Massachusetts General Hospital, and a colleague. In an editorial in Radiology, they warn that radiologists who limit their activities to image interpretation will become a commodity and may eventually become obsolete.

The editorialists predict a variety of changes in the science and practice of radiology, urging colleagues to search out and demonstrate new ways of adding value in clinical care.

Radiology Workflow and Settings

Workflows will need to be driven by disease focus rather than patient location:

  • Some inpatient care is now being delivered at home, suggesting more urgent and acute care will be delivered at home as well
  • Portable equipment will become available for radiography, motionless CT, and low-field-strength MRI
  • Teleradiology will become even more common due to the need for 24/7 coverage, outreach beyond urban centers, and flexible physician work hours

Artificial Intelligence (AI)

The greatest risk to radiology is that referring physicians with clinical information at their fingertips will use AI for independent image interpretation. This trend may extend to patients themselves. For example, work is well underway to facilitate simple ultrasound data acquisition with AI-powered smartphones.

Continued improvements in AI will change all aspects of radiology practice:

  • Algorithm development will broaden to include a comprehensive evaluation of all possible features present in certain imaging examinations
  • Algorithms will become more robust and will be maintained with "lifelong learning"; radiology trainees will require a basic understanding of algorithm design and training so they can be the keepers of these algorithms and supervise their use
  • AI will affect all aspects of clinical practice, from triage to lesion characterization; as imaging findings are integrated with other clinical indicators, radiologists should participate in assessing disease and predicting treatment outcome

Precision Imaging and Image-guided Intervention

The use of non–image-based precision diagnostics (e.g., "liquid biopsies") and minimally invasive interventions in other specialties is certain to grow. Percutaneous endoscopic imaging will enable interventional radiologists to perform minimally invasive interventions in domains previously reserved for other specialties.

Imaging centers will evolve beyond diagnostic centers to become treatment planning and prediction centers. Radiologists will be responsible for phenotyping with imaging markers that predict treatment response on a large scale.

Subspecialized training will remain extremely important because radiologists' clinical relevance can be assured only if they are as knowledgeable as subspecialized referring physicians. Likewise, it will be critical for radiology to partner with other disciplines—such as medicine, data science, and engineering—to perform clinically relevant research and carve out areas for innovation and expertise.


Radiotheranostics combines targeted diagnostic imaging—currently with PET or SPECT—and radionuclide therapies. Real-time feedback from imaging allows clinicians to improve patient selection for therapy and assess response sooner and more precisely.

Oncologic radiotheranostics has enormous untapped potential to treat a range of additional cancers, given the capacity of radionuclide probes to be adapted to different targets and treat specific cells. The probes also have the potential to be applied to other diseases where precision medicine is possible, including certain inflammatory and autoimmune conditions.

Learn more about the Department of Radiology

Learn more about research in the Department of Radiology


Manisha Bahl, MD, MPH, a radiologist in the Breast Imaging Division in the Department of Radiology, explains the importance of a recent systematic review and meta-analysis that includes all literature published to date on contrast-enhanced mammography.


Daniel B. Kopans, MD, of the Radiology Department at Mass General, says it's time to stop researching digital breast tomosynthesis using sequential cohorts. He suggests a simple, direct method to study whether replacing synthetic two-dimensional images with full-field digital mammographic images is warranted.