![\"Chart](\"https:\/\/rtmedical.com.br\/wp-content\/uploads\/2026\/04\/volume-exames-radiologia-queda.jpg\")
The authors’ central argument is that an alleged “imaging slowdown” may have multiple causes. Their hypotheses include physicians increasingly recognizing the potential harms of unnecessary scans, the impact of the Choosing Wisely<\/em> campaign, reimbursement changes, prior authorization requirements, accountable care organizations, and efforts to halt unnecessary repeat exams.<\/p>\n Cutler and Klarnet acknowledge that “there is no consensus on which of these hypotheses explain imaging trends, or whether there are other important factors.” Cutler was part of a Health Policy Commission created to help reduce medical spending in Massachusetts, which provides context for his cost-focused perspective.<\/p>\n One of the article’s most controversial proposals is the suggestion that artificial intelligence could help reshape radiological practice. Cutler describes analyzing medical images as a “straightforward process for machine learning,” stating that computer programs can already “emulate radiologists” in certain scenarios. However, the authors themselves caution that “it is unlikely that computers will replace radiologists any time soon” and that “there are many situations where human ability outperforms AI.”<\/p>\n Their proposal includes a triage model where radiologists would read the most complex exams regardless of the patient’s geographic location, while AI programs would handle routine cases<\/a>. In rural areas where radiologists are scarce, the authors suggest that a computerized diagnosis could be “a reasonable substitute” for a trained radiologist.<\/p>\n Cutler’s thesis faces significant objections from practicing professionals. Many radiologists argue that when imaging volume data is adjusted for complexity and modality, it does not support a narrative of generalized slowdown. Modalities such as MRI and PET-CT continue to expand, and the growing adoption of population-based screening programs, such as mammography and low-dose CT for lung cancer<\/a>, is expected to drive future demand.<\/p>\n Another point of criticism is the premise that AI can serve as a substitute in areas with professional shortages. Medical societies such as the ACR (American College of Radiology) have been monitoring regulatory proposals<\/a> related to AI in radiology, emphasizing that the technology should be viewed as a support tool, not a replacement for medical judgment.<\/p>\n The analysis deserves additional context. Data from the FDA shows that radiology leads among specialties with approved AI devices<\/a>, indicating a trend toward integration rather than replacement. Furthermore, market reports project consistent growth in the diagnostic imaging sector over the coming decade, driven by population aging, expanded access, and new clinical indications.<\/p>\n The discussion about exam volumes must also account for the fact that study complexity has increased significantly. A single CT scan can generate hundreds of images requiring detailed analysis, and advanced techniques such as perfusion, diffusion, and spectroscopy add layers of information that demand more time per exam. Reducing the analysis to “number of exams” ignores this reality.<\/p>\n While the U.S. debate focuses on potential over-utilization, much of the world faces the opposite challenge. Countries across Latin America, Africa, and Asia still struggle with extensive imaging backlogs, unequal coverage between urban and rural areas, and radiologist shortages that limit access to timely diagnosis. The American discussion about “excess” imaging has limited relevance in these contexts, where the challenge is expanding access with quality.<\/p>\n That said, the portion of the article about AI as a support tool in remote areas resonates with existing teleradiology initiatives that allow specialists in major centers to read exams performed in distant communities. In this model, AI could serve as a triage and prioritization layer, accelerating turnaround times without replacing the radiologist.<\/p>\n The article by Cutler and Klarnet contributes to a necessary debate, but its conclusions should be read with caution. An economist’s cost-containment perspective is valuable but incomplete without considering the growing complexity of exams, emerging clinical indications, and unmet demand in developing countries. The future of radiology lies in intelligent integration of AI into the workflow, not in reducing the radiologist’s role.<\/p>\n Source:<\/strong> JAMA Health Forum<\/a> via Health Imaging \/ Radiology Business<\/a><\/p>\n","protected":false},"excerpt":{"rendered":" JAMA Health Forum article argues imaging volume growth is slowing and AI could reshape radiology practice.<\/p>\n","protected":false},"author":1,"featured_media":17293,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"ngg_post_thumbnail":0,"fifu_image_url":"","fifu_image_alt":"","footnotes":""},"categories":[113,112],"tags":[],"class_list":{"0":"post-17308","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-inteligencia-artificial","8":"category-radiologia"},"aioseo_notices":[],"rt_seo":{"title":"","description":"Harvard economist argues imaging volumes are declining and AI could reshape radiology. Critical analysis and global implications.","canonical":"","og_image":"","robots":"index,follow","schema_type":"Article","include_in_llms":true,"llms_label":"Imaging Volumes and AI in Radiology","llms_summary":"Harvard article in JAMA Health Forum argues imaging volume growth is slowing and proposes AI as solution, sparking debate across the radiology community.","faq_items":[],"video":[],"gtin":"","mpn":"","brand":"","aggregate_rating":[]},"_links":{"self":[{"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/posts\/17308\/"}],"collection":[{"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/posts\/"}],"about":[{"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/types\/post\/"}],"author":[{"embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/users\/1\/"}],"replies":[{"embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/comments\/?post=17308"}],"version-history":[{"count":1,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/posts\/17308\/revisions\/"}],"predecessor-version":[{"id":17310,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/posts\/17308\/revisions\/17310\/"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/media\/17293\/"}],"wp:attachment":[{"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/media\/?parent=17308"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/categories\/?post=17308"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/tags\/?post=17308"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}AI as a Path Forward for Radiology<\/h2>\n
Pushback from the Radiology Community<\/h2>\n
What the Data Actually Shows<\/h2>\n
Global Implications<\/h2>\n
Future Outlook<\/h2>\n