|
AI’s ROI Paradox, FDA Rejects AI Petition, and Reports Scare Patients
April 13, 2026
|
|
|
|
|
Together with
|
|
|
|
“If you work from a shared reporting list and only pick the easy ankle MRIs, you’re not a master radiologist — you’re just a cherry-picker hiding in the reading room. Confident radiologists need to tackle the hardest cases head-on. You are a specialist — you’re not a resident anymore.”
|
|
Christoph Agten, MD.
|
|
|
The future of imaging will be heavily influenced by artificial intelligence, both in radiology and pathology, and this convergence will have major implications for population health. We talked to Razik Yousfi and Chiranjiv Singh of Tempus AI about this exciting future in this edition of The Imaging Wire Show.
|
|
|
|
As radiology AI slowly moves from pilot projects to widespread clinical adoption, a new survey reveals a paradox: The technology is popular with radiologists, but few imaging facilities using AI have collected hard data showing its return on investment.
AI’s slow clinical adoption has frustrated both clinicians and algorithm developers alike, but the technology is gaining steam.
- Despite growing clinical evidence, research on AI’s financial value and ROI has been slower in coming.
To remedy that situation, AI governance startup Croviz.ai conducted a study of 445 radiology AI users on the economics and evaluation of AI. The full report is available here.
- Survey respondents came from 12 different countries and included a variety of professional roles, including vendor executives, radiologists, and IT and informatics personnel.
Croviz founders Ayman Talkani and AadilMehdi Sanchawala found that while radiology AI power users loved the technology – and some refused to work without it – few had determined a positive financial return from it. Findings included…
- 95% of sites already using AI had renewed at least one contract with an AI vendor in the last 12 months.
- But only 30% had quantified a positive financial ROI from AI.
- 54% cited better quality of life for radiologists as their main reason for renewing an AI contract.
So if AI’s value hasn’t been demonstrated, why are radiology sites renewing AI contracts?
- The number one reason cited by 54% of those renewing contracts was because their radiologists felt AI improved their quality of life – the only outcome measure leadership could quickly measure with qualitative user feedback.
- Lower on the scale was reduced turnaround time (18%), more scans per reader (10%), reduced downstream patient costs (10%), and better diagnostic accuracy (8%).
- Just 6% paid attention to hard metrics like staff retention rates.
What’s the best way out of the AI ROI paradox? The Croviz researchers recommended more frequent and transparent AI governance.
- Survey respondents who monitored AI performance more closely – such as more often than once per quarter – exhibited more trust in AI.
The Takeaway
The new survey offers an intriguing look at AI adoption and the question of ROI for the technology. It suggests that – much like another digital technology, PACS – AI adoption is being driven more by its popularity among radiologists than hard ROI considerations.
|
|
The Power of a Smooth Go-Live
Don’t gamble on your healthcare institution’s go-live: take control of your PACS migration with ENDEX from Enlitic. Discover how ENDEX uses AI to standardize, normalize, and cleanse your imaging metadata before migration.
|
|
A Comprehensive Suite of Tools in a Single Platform
UnityVue from Mach7 Technologies offers a comprehensive suite of tools in a single platform, streamlining the entire radiology process from image acquisition to reporting. Discover what it can do for you today.
|
|
Intelligence and Quality in Bedside Imaging
The new FDR Go iQ from Fujifilm Healthcare Americas is one of the smallest, quietest, and most agile full-size portable X-ray systems available. Together with Fujifilm’s trademark image quality and dose performance, FDR Go iQ brings premium imaging to bedside exams.
|
|
- FDA Rejects Harrison AI Petition: The FDA rejected a petition by Harrison.ai to change how it regulates algorithms to make it easier for developers to get clearance for multiple AI models. The petition suggested an alternative regulatory path in which once an AI firm got an FDA clearance, it could comply with post-market requirements rather than submit regulatory applications for every new product. The agency’s denial is a blow to AI firms that focus on multi-abnormality detection, as they will continue to have to draw up new applications for every new algorithm.
- Radiology Reports Scare Patients: A new study underscores how radiology reports written for clinicians can scare patients. Researchers showed a group of 20 people a sample report with a coronary artery calcium score of 0 and an incidental lung nodule finding. Some 85% said the report confused them and 40% said they would experience fear if the report were their own. Major sources of confusion were the CAC scoring system, medical jargon, and acronyms, and 95% said they would appreciate a lay summary with fewer medical terms.
- Topol Supports Mammography AI Review: Mammography AI just won an influential new ally in cardiologist and medical blogger Eric Topol, MD, who just published an editorial in The Lancet advocating for AI review of all breast screening exams. Topol cites evidence supporting AI from trials like MASAI, as well as recent studies on AI risk prediction. “Efforts are needed across different types of health systems to accelerate implementation of AI-supported mammography,” he wrote.
- Simpler Radiology Reports with LLMs: Could large language model AI tools be used to produce patient-friendly report summaries without bogging down radiology personnel? Researchers tried it in a new study in JACR, finding that GPT-o1 produced summaries of 30 CT reports in multiple languages that radiologists rated as acceptable (median score of 4.0 out of 5), with 77% rated as “extremely accurate” or “very accurate.” GPT-o1 also performed better than Google Translate for translation accuracy (4.0 vs. 2.0-3.0).
- Point-Counterpoint on Photon-Counting CT: With vendor competition heating up in photon-counting CT, AJR last week published point-counterpoint articles on the technology. On the pro side, Giuseppe Toia, MD, notes that PCCT doubles spatial resolution and image contrast, enabling better visualization of potentially cancerous lesions. On the con side, Marc Zins, MD, maintains there is insufficient evidence to support PCCT’s impact on accuracy and patient outcomes, particularly for abdominal imaging. PCCT scanners also produce large image sets, which could complicate workflow, especially in oncology imaging.
- Philips Gets Verida CT Clearance: The FDA cleared Philips’ Verida, a detector-based spectral CT scanner the company introduced at RSNA 2025. In developing Verida, Philips integrated AI across the imaging chain, from acquisition to reconstruction, improving spectral image quality with lower noise and enabling up to 45% radiation dose reduction. Philips will emphasize the advantages of detector-based spectral CT compared to techniques that require dual X-ray sources or kVp switching. The authorization comes after the FDA cleared the company’s Rembra CT family last month.
- Radiation Dose in Pediatric ‘Frequent Flyers’: There’s good news and bad news in a new study in Pediatric Radiology on pediatric CT radiation dose. The good news is that in a review of 10.2k children scanned from 2013 to 2023, 89% only got 1-2 scans, and just a small percentage exceeded cumulative effective dose guidelines by age 16 (0.95% over 50 mSv and 0.2% over 100 mSv). The bad news is that’s still a lot of radiation for a young person, highlighting the need for continued vigilance around radiation dose.
- Opportunistic Head CT for CVD Risk: Most AI tools for opportunistic screening of cardiovascular disease risk use thoracic chest CT scans, but a new study in JACC instead used head CT exams. Researchers estimated CVD risk and coronary artery calcification in 28k patients using vision transformer-based deep learning models to analyze head CT. A model that used CT head scans added to PREVENT-CVD guidelines had higher predictive accuracy for CVD risk than PREVENT-CVD alone (0.82 vs. 0.75). For CAC estimation, the head CT model was slightly better (0.76 vs. 0.75).
- AI Predicts Heart Failure from Visceral Fat: In another study in JACC, researchers used a homegrown deep learning radiomics model to predict heart failure from analysis of visceral fat (epicardial adipose tissue) from CT scans. U.K. researchers tested their fat radiomic profile for heart failure in 13.4k patients, finding that over a follow-up of at least four years, each 25 percentile increase in FRPHF led to a nearly 4X higher heart failure risk. Patients in the highest risk decile had nearly 20X the risk as those in the lowest.
- Brain Swelling Linked to Worse Stroke Outcomes: Head CT scans showed that stroke patients with brain swelling had worse outcomes in a new meta-analysis in Stroke. Researchers analyzed nine studies covering 1.5k patients, finding that signs of perihematomal edema on CT within 24 and 72 hours after intracerebral hemorrhage onset were linked to worse outcomes. Every 1 mL of edema was linked to a 4% higher risk of death or dependence. The findings illustrate the importance of treating stroke patients for secondary brain injury like edema.
- How Common Are Safety Events in Radiology? How often do safety events occur in radiology? A new study in Current Problems in Diagnostic Radiology sheds some light, finding that in two years at a single academic medical system, radiology accounted for 3.2% of 22.3k events. Most radiology events involved medication and contrast administration (23%) and communication and coordination failures (22%), followed by patient handling and physical safety (19%). The good news is that 72% of events caused no emotional distress or patient harm.
- GE’s Ultrasound Integration: GE HealthCare made improvements to its bkActiv intraoperative ultrasound scanner that improve its utility for surgical navigation. bkActiv now features digital integration with Medtronic’s Stealth AXiS surgical navigation system, bringing planning, navigation, and robotics into a single platform for use during cranial procedures.
- Outreach Boosts Post-COVID Screening: In a study that provides lessons for any facility looking to improve screening rates, researchers from UMass Memorial Medical Center were able to boost mammography rates following the COVID-19 pandemic. In a new paper in Surgery, they describe how interventions like redesigned waiting areas, better scheduling, a marketing campaign, and nurse navigators helped them counter a 16% drop during the pandemic era to achieve an 80% increase compared to April 2021. Time to appointments also improved.
- SHINE Moves Isotope Production Ahead: Isotope developer SHINE got a big boost in its quest to create a U.S. source for medical isotopes, securing commitment for a $263M loan from the U.S. Department of Energy. Once approved, SHINE will use the funding to complete Chrysalis, a medical isotope production facility it is building to create isotopes like molybdenum-99, the precursor to technetium-99m, the most commonly used radioisotope for nuclear cardiology. The U.S. relies on imports for its Mo-99 supply.
|
|
Stay in Flow
Radiologists shouldn’t think about their reporting software while interpreting exams. Workflow should protect concentration – not interrupt it with clicks and friction. See how Rad AI Reporting keeps the focus on the read. Book a demo.
|
|
The Advantages of Cloud-Based PACS
CloudPACS offers significant advantages to healthcare providers for enterprise imaging, including no on-premise requirements and enhanced reliability. Learn more about the advantages of CloudPACS in this talk by Visage Imaging’s Steve Deaton.
|
|
The Workstation of the Future
A dedicated team of radiologists shapes every aspect of the functionality and design of MosaicOS from Mosaic Clinical Technologies. Learn how every feature was built to eliminate distractions, amplify focus, and enhance the radiologist experience.
|
|
- Experience the Breakthrough in 3T MRI Innovation: Philips empowers you to advance clinical and diagnostic confidence with cutting-edge 3T MRI technology to deliver precision imaging for every patient, every exam, every day. Learn more today.
- Advanced Medical Imaging in Small-Town Iowa: Clarinda Regional Health Center has been a cornerstone of care in Clarinda, Iowa, since opening its doors in 1939. Learn how CRHC partnered with United Imaging to bring advanced imaging capabilities directly to its patients.
- Covering All the Bases for Pulmonary Nodule Detection: Improving the detection of missed nodules means better patient outcomes. Learn more about ClearRead Chest from Riverain Technologies – the fully automatic chest AI for CT and X-rays that improves detection and efficiency – and be sure to visit them at RBMA 2026.
- AI-Centric Radiology Reporting: KailoAir is a new AI-centric radiology reporting solution from Kailo Medical that combines real-time voice dictation, AI-powered prior study analysis, and structured reporting in a seamless browser-based workspace. No installation and no compromise.
- Redefining Breast Imaging in the Enterprise Era: As breast imaging grows more complex, radiology teams need more than siloed tools. Watch this on-demand webinar hosted by AGFA HealthCare about transforming breast imaging workflows in the enterprise imaging era.
- Diagnosing Cardiac Amyloidosis with Echo AI: Diagnosing cardiac amyloidosis on echocardiography can be challenging. In a new study, echo AI algorithms from Us2.ai were used to improve an existing multiparametric CA scoring system, and also perform automated video-based amyloidosis detection.
- Building an Enterprise Imaging Ecosystem: MultiCare Health System wanted to create an enterprise imaging foundation to secure its own future and that of smaller clinics and rural hospitals that need support providing patient care. Find out how they did it with Merge in this case study.
- Radiology Case Report: A man in his 50s presented with syncope with minor head trauma and unassociated risk factors. Find out how MRI helped provide a diagnosis in this case study.
- Radiology Automation Simplified: CARPL is an enterprise-grade radiology AI validation and deployment platform with 250+ AI applications across 85+ AI vendors that empowers healthcare providers to access, assess, and securely integrate imaging AI in their practice. Book a demo today.
- The World’s Happiest Customers – 13 Straight Years: Sectra continues to top the charts when it comes to customer satisfaction. The company recently won seven Best in KLAS awards, including the 13th straight for its Sectra PACS enterprise imaging solution. Find out why on this page.
- Experience Dynamic Simplicity in Fluoroscopy: Still on the fence about upgrading to the latest technology that fluoroscopy has to offer? Hear from your peers about their experiences with the first-ever installed LUMINOS Q.namix R system from Siemens Healthineers.
- Leveraging AI-Powered Discovery for Image Exchange: Southwest Medical Imaging, a premier physician-owned radiology practice in the southwestern U.S., partnered with Medicom to streamline their workflow. Discover how they utilize Medicom’s AI-powered Smart Search, which leverages an LLM to automatically detect and surface imaging data, to eliminate manual searching and accelerate patient care.
- Transform Imaging Data into Actionable Predictions: When you choose Quibim, you get more than a partner for detecting and diagnosing prostate cancer on MRI scans. Learn how they can help you transform imaging data into actionable predictions by booking a demo today.
- Remote Reading Is Essential for Enterprise Imaging: Remote reading has become essential to enterprise imaging operations because healthcare organizations need scalable, flexible solutions. Discover how cloud-based solutions from Intelerad can help solve remote reading challenges.
- Elevating Breast Cancer Detection: Breast Suite from DeepHealth is a new package of AI-powered solutions delivering increased breast cancer detection rates, risk stratification tools, and viewing and reporting workflow acceleration. Find out how it can benefit your practice today.
- Cardiac CT Online Training Course: Medality’s online cardiac CT training course is designed for busy clinicians working towards Level 2 Cardiac CT certification. Endorsed by SCCT, it provides a flexible, practice-focused approach to develop your coronary CTA interpretation and reporting skills while earning CME.
|
|
|
|
|
