|
Malpractice Reform, Low-Value Imaging, and Rad Subspecialization
October 30, 2025
|
|
|
|
|
Together with
|
|
|
|
“The recent leaps in generative AI have changed what’s possible. What if an assistant could read the chart end to end, integrate imaging results, listen in on the consultation, and then whisper the key facts and insights the clinician needs right now?”
|
|
Chris Mansi, MD, and Andrew Ibrahim, MD, of Viz.ai in NEJM AI.
|
|
|
Artificial intelligence looks set to have the biggest impact on healthcare since the arrival of PACS and EHR in the 1990s. But is healthcare about to make the same mistakes with AI deployment that it made back then? Learn about what healthcare can do differently to avoid the pitfalls of the past in this contribution from Densitas.
|
|
|
|
We all know it happens – medical imaging scans of questionable clinical value, performed not to improve patient diagnosis but to defend clinicians in the event of malpractice litigation. A new study in AJR supports the idea that defensive medicine is driving up imaging use by finding a link between malpractice reform and lower emergency imaging utilization.
The proliferation of imaging technology throughout the healthcare enterprise – and especially in the emergency setting – gives clinicians a powerful tool that’s just too tempting not to use.
- Head CT scans can quickly rule out patients who might have a hemorrhagic stroke, for example, while cardiac CT angiography is showing its value for working up patients with chest pain.
But with great power comes great responsibility. Unnecessary imaging not only drives up healthcare costs but can expose patients to additional radiation as well as complications from working up suspicious findings.
- Medical-legal experts speculate that malpractice reform through tools such as damage caps could tamp down defensive medicine by limiting physicians’ legal exposure to lawsuits in the event they make a mistake.
In the new study, researchers from the ACR’s Harvey L. Neiman Health Policy Institute tested the idea by analyzing 630k Medicaid encounters for patients with headache presenting to the emergency department in 2019.
- They then correlated head and neck imaging volume to various factors that could influence utilization, including whether states had implemented tort reform.
Their analysis discovered that emergency imaging utilization was less likely to occur…
- In states with laws on “several liability” (in which parties are only responsible for their own share of damages) (OR = 0.68).
- In states with malpractice damage caps (OR = 0.79).
- In states with greater mean malpractice payment (although the effect size was minimal; OR = 0.99).
A couple other interesting findings included…
- Referring physicians other than emergency medicine were far more likely to order more imaging (OR = 8.45).
- Facilities with fewer than 100 beds were less likely to order imaging (OR = 0.65).
The Takeaway
The new findings linking malpractice reforms with lower emergency imaging use confirm what many of us have already suspected. Whether they lead to health policy reforms remains to be seen.
|
|
Reimagining Radiology with Apple Vision Pro
Discover how Apple Vision Pro is helping to pioneer what’s possible in radiology in this special event in Chicago during RSNA 2025. Hear from key opinion leaders and Visage Imaging executives on how spatial computing is transforming radiology.
|
|
Experience the Future of Radiology
Visit Enlitic at RSNA 2025 booth #7730 to experience the future of radiology and see how the company is revolutionizing radiology data migration with its Laitek acquisition. Schedule a meeting today.
|
|
Tools for Lung Cancer Screening in Europe
As lung cancer screening programs gear up to launch across various European countries, the integration of AI nodule detection tools promises to enhance the accuracy of low-dose CT scans. Watch this video from Riverain Technologies to learn more.
|
|
- Can AI Find Unnecessary Imaging? Could AI be used to reduce unnecessary medical procedures like advanced imaging scans? CMS plans to find out with WISeR (Wasteful and Inappropriate Service Reduction), a pilot project in six U.S. states that will leverage AI to detect low-value services and assess Medicare coverage determinations for certain services. Imaging isn’t specifically mentioned on the agency’s information page about WISeR, but it wouldn’t be surprising if certain exams (like CT for headache) fell under its purview.
- Subspecialization Linked to Practice Closures: A new analysis in JACR found that radiologists leaving a practice that just closed were 10% more likely to be practicing as subspecialists a year later. In an analysis of 239k radiologist-years from 2014-2022, the percentage of subspecialized radiologists grew (46% to 57%), with the rate of subspecialization growing faster among those with a recent practice closure (+2.5% vs. +1.8%). Practice closures could be driving radiologists to larger groups that support increased subspecialization.
- AI in Radiology: New Problems, Old Tech: AI’s success in radiology won’t be defined by elegance of algorithms but by pragmatism of deployment. There are ways for imaging practices to adopt AI and then evolve their processes to make it work. Find out what they are in this contribution from Densitas.
- Lower Radiation with Cardiac MR Workup: Using cardiac MRI rather than invasive angiography to work up emergency patients with chest pain resulted in lower overall radiation dose, including from downstream exams. In the CMR IMPACT trial, researchers evaluated both strategies in 312 patients, finding that CMR workup generated 40% lower overall radiation dose at the initial patient visit (4.7 vs. 7.8 mSv) and 21% lower at one year (7.5 vs. 9.5 mSv). But limited availability of cardiac MR in the emergency setting could limit wider CMR adoption.
- Cardiovascular Aging’s Link to Brain Health: Another research study is linking subclinical measures of cardiovascular disease risk to reduced brain function on MRI scans, in some cases decades later. Researchers followed 1.4k patients from the MESA study for 17-20 years, finding that their baseline composite scores for subclinical arteriosclerosis and atherosclerosis correlated with brain MRI measures acquired later, such as reduced gray matter and greater burden of white matter hyperintensities. The findings illustrate the link between cardiovascular aging and general brain health later in life.
- Mirai Predicts Future Interval Cancer Risk: MGH’s Mirai deep-learning algorithm predicted risk of interval breast cancer over three years in a new U.K. study published in Radiology. Mirai analyzed normal mammograms from 134.2k women who were screened every three years, generating risk scores for future breast cancer. AUC scores varied slightly by years of follow up (one = 0.72, two and three = 0.67), showing that Mirai could help determine which at-risk women might need additional screening in between three-year rounds.
- Why Does AI Miss Breast Cancer? We know breast cancer AI isn’t perfect, but we don’t know why it sometimes misses cancers. Researchers from South Korea investigated the question in a secondary analysis of the AI-STREAM trial in Radiology: Artificial Intelligence. Lunit’s Insight MMG algorithm analyzed mammograms from 24.5k women, and by itself had 90% sensitivity and 94% specificity. False negatives were more common in dense breasts, and 80% of AI’s misses were detected by radiologists during screening (having access to prior images helped humans).
- NVIDIA Updates Clara AI Tools: NVIDIA this week updated its Clara family of AI tools for healthcare and scientific research, with the new launches including Clara Reason, an open-source vision language model for what the company calls “chain-of-thought reasoning” to advance explainable AI medical research. Clara Reason and other NVIDIA tools are available on major open-source platforms.
- Philips Adds CT/Cath Lab Integration: At this week’s TCT 2025, Philips announced a new feature for its workstation software that brings pre-operative CT images directly into cath lab workflow. The company has integrated its Advanced Visualization Workspace software with the Azurion image-guided therapy system through a new capability called Follow C-arm that automatically synchronizes 3D reconstruction of coronary arteries with Azurion C-arm’s movement. The feature is the first step toward CT-guided percutaneous coronary intervention.
- AI Plaque Analysis Predicts Adverse Events: In other news from TCT 2025, researchers presented the latest results from the CONFIRM2 trial with Cleerly’s AI software for quantitative coronary CT analysis. In 6.6k symptomatic patients, AI-QCT showed that noncalcified plaque volume and stenosis diameter predicted heart attack and death, with patients in the highest third of noncalcified plaque volume having nearly twice the risk (HR = 1.93). Also, many adverse events happened in patients without obstructive coronary artery disease.
- Elucid Adds Carotid to Plaque AI: Elucid is launching a new indication for its PlaqueIQ solution for identifying arterial plaque on CT scans. PlaqueIQ was already being marketed for coronary plaque, and Elucid last month received FDA clearance to sell the solution for carotid plaque as a tool for identifying patients at risk of stroke. PlaqueIQ performs quantitative and qualitative assessment of atherosclerotic risk and can help identify lesions at high risk of rupture.
- Plaque Analysis AI Gets FDA Nod: Another new tool for coronary plaque analysis using CT scans has been cleared by the FDA, this one from Circle Cardiovascular Imaging. The company’s cvi42 | Plaque solution quantitatively assesses coronary plaque and identifies patients at risk of cardiovascular disease. The software is integrated into users’ CT workflow and performs automated segmentation of the coronary lumen and wall, as well as per-lesion and per-vessel analysis.
- Viz.ai AI Agents for Patient Summaries: Viz.ai launched Viz Assist, a suite of new generative AI agents that create patient summaries from EHR data and ambient listening that can be combined with data from the company’s AI algorithms. Viz Assist supports providers by delivering clinical insights, scanning proactively for signals that indicate patient changes and alerting clinicians in real time. The launch represents a major expansion for Viz, from its core in medical imaging and ECG AI to the broader patient management segment.
- QT Imaging Deploys Intelerad PACS: Breast imaging developer QT Imaging will deploy Intelerad’s PACS and image sharing solutions across its clinical research and commercial networks. Intelerad’s ImageShare cloud-based technology will be integrated into QT Imaging’s Breast Acoustic CT scanners to enable QT customers to upload, store, anonymize, and share imaging data efficiently. The agreement includes InteleShare Research PACS and InteleShare Cloud PACS.
- Presidents Get Procedures: Current and former U.S. presidents are in the news this week in the world of diagnostic and therapeutic radiology. President Donald Trump disclosed on Monday that he received an MRI scan during a visit earlier this month to Walter Reed National Military Medical Center – Trump said the results were “perfect.” Meanwhile, former President Joe Biden completed a course of radiation therapy at the University of Pennsylvania as part of treatment for prostate cancer.
|
|
What Would You Do with an Extra Hour?
Rad AI Impressions automatically generates your report impressions in your own voice, from your dictated findings. Radiology teams already use Rad AI to give radiologists their time back. See what you could do with yours. Book your RSNA meeting today.
|
|
A New Direction at Mach7
Enterprise imaging company Mach7 Technologies has embarked on a new direction with recently appointed CEO Teri Thomas at the helm. Learn about the new focus in this Imaging Wire Show.
|
|
- Gain Clarity in MRI at Speed: Discover how STAGE from SpinTech MRI gives you better gray-white matter contrast in MRI and more efficient reads, with up to 30% faster scans on all 1.5T and 3T magnets.
- Innovation for the Imaging Workflow You Deserve: Radiologists are overwhelmed by complex systems and growing caseloads. Discover how innovative workflow orchestration from Merge integrates various PACS, RIS, and EMR systems into a single AI-driven workspace.
- Abdominal Ultrasound Worksheets and Reports in 45 Seconds: Visit Kailo Medical at booth #4555 at RSNA 2025 to discover augmented reporting that produces worksheets and reports in 45 seconds, improving accuracy, boosting efficiency, and delivering better outcomes for patients and providers alike.
- Digital Radiography Designed for Freedom: United Imaging’s new uDR Aurora CX digital radiography system was designed to give you unprecedented freedom in imaging patients thanks to its new advances in computer vision, automation, and AI diagnostics.
- Reimagine Imaging Innovation: Experience a new era of diagnostic imaging in the AGFA booth #2565 at RSNA 2025 in which every click, case, and connection is designed around how you work. Request an on-site demo today.
- Advanced AI for Prostate MRI: QP-Prostate from Quibim is your advanced solution for detection and diagnosis of prostate cancer from MRI scans. Discover how it streamlines your workflow by detecting suspicious lesions, segmenting the prostate, and ensuring compliance with PI-RADS V2.1 guidelines.
- Presenting Unboxing AI: Check out CARPL’s video series, Unboxing AI, featuring experts discussing AI and its future in radiology. The next episode on October 30 features Ashwin Amarapur of Aikenist Technologies – reserve your seat today.
- Remote vs. Tableside Fluoroscopy: Choosing the right fluoroscopy system is critical – download this e-book from Siemens Healthineers to discover key safety and efficiency factors that will help you decide between remote and tableside options.
- A New Standard of AI-Powered Care: Visit DeepHealth at RSNA 2025 booth #1329 to learn how their integrated, cloud-native solutions can unify your imaging experience. Book a meeting today.
- Echo AI for Cardiac Amyloidosis: AI is reshaping the way cardiac amyloidosis is detected. A notable example is Us2.ai’s pattern recognition model, which can identify the disease from a single apical four‑chamber echo view.
- Unlock Next-Generation AI with Foundation Models: Learn about Microsoft’s new family of cutting-edge multimodal medical imaging foundation models designed for healthcare organizations to test, fine-tune, and build tailored AI solutions specific to their needs, while minimizing extensive compute and data requirements.
- Discover What’s Next in Radiology at RSNA 2025: Join Philips at RSNA 2025 in booth #6730 to see revolutionary new innovations in MRI, CT, and informatics. Don’t miss your chance to witness breakthroughs that will redefine what’s possible at 10:30 am on Sunday November 30.
- Radiology Case Report: A female in her 40s without symptoms presented for screening mammography. Discover how information from contrast-enhanced MRI led to a finding of carcinoma.
- Building the Future of Radiology: Visit Medality at RSNA 2025 booth #1140 to learn how they are at the forefront of radiology practice growth and training as the simulated learning platform that everyone loves. Book a demo today.
- Every Image Tells a Story at RSNA 2025: Visit Intelerad at RSNA 2025 booth #6113 to learn how their solutions are empowering radiologists and patients while transforming radiology workflows. Schedule a meeting today.
- It’s Time to Make AI Adoption Simple: Visit Gleamer at RSNA 2025 booth #4936 to experience a milestone moment that will reshape AI’s role in medical imaging. Expect more than a sneak peak: Expect the future.
|
|
|
|
|
