Larsen M et al, writing in Radiology: Artificial Intelligence. 

Imaging Wire Sponsors

Bayer  •  Blackford  •  CARPL.ai  •  Clearpath  •  DeepHealth  •  Enlitic  •  Gleamer  •  Intelerad  •  Medality  •  Merge by Merative  •  Nuance  •  PocketHealth  •  Riverain Technologies  •  Siemens Healthineers  •  SpinTech MRI  •  TeraRecon  •  United Imaging  •  Us2.ai  •  Visage Imaging

Artificial Intelligence

Fine-Tuning AI for Breast Screening

AI has shown in research studies it can help radiologists interpret breast screening exams, but for routine clinical use many questions remain about the optimal AI parameters to catch the most cancers while generating the fewest callbacks. Fortunately, a massive new study out of Norway in Radiology: Artificial Intelligence provides some guidance. 

Recent research such as the MASAI trial has already demonstrated that AI can help reduce the number of screening mammograms radiologists have to review, and for many low-risk cases eliminate the need for double-reading, which is commonplace in Europe. 

• But growing interest in breast screening AI is tempered by the field’s experience with computer-aided detection, which was introduced over 20 years ago but generated many false alarms that slowed radiologists down. 

Fast forward to 2024. The new generation of breast AI algorithms seems to have addressed CAD’s shortcomings, but it’s still not clear exactly how they can best be used. 

• Researchers from Norway’s national breast screening program tested one mammography AI tool – Lunit’s Insight MMG – in a study with data obtained from 662k women screened with 2D mammography from 2004 to 2018. 

Researchers tested AI with a variety of specificity and sensitivity settings based on AI risk scores; in one scenario, 50% of the highest risk scores were classified as positive for cancer, while in another that threshold was set to 10%. The group found …

• At the 50% cutoff, AI would correctly identify 99% of screen-detected cancers and 85% of interval cancers. 
• At the 10% cutoff, AI would detect 92% of screen-detected cancers and 45% of interval cancers 
• AI understandably performed better in identifying false-positive cases as negative at the 10% threshold than 50% (69% vs. 17%)
• AI had a higher AUC than double-reading for screen-detected cancers (0.97 vs. 0.88)

How generalizable is the study? It’s worth noting that the research relied on AI of 2D mammography, which is prevalent in Europe (most mammography in the US employs DBT). In fact, Lunit is targeting the US with its recently cleared Insight DBT algorithm rather than Insight MMG. 

The Takeaway

As with MASAI, the new study offers an exciting look at AI’s potential for breast screening. Ultimately, it may turn out that there’s no single sensitivity and specificity threshold at which mammography AI should be set; instead, each breast imaging facility might choose the parameters they feel best suit the characteristics of their radiologists and patient population. 

A New Solution for Cardiac MR Programs

Intelerad is making advanced cardiac solutions available whenever you want them. Learn more about the company’s new Precession application, which enables you to view, analyze, and report cardiac MR exams – all within a standard web browser.

4 Reasons Why Subscription Is Better

The KLAS-leading Intuition advanced visualization solution from TeraRecon includes all the clinical features you need, plus a growing list of exclusive subscription-only content. Learn more about what’s included.

The Wire

• Dedicated Neuro MRI Scanner: In a prize-winning presentation at ISMRM 2024, researchers reported on their initial experiences with a 0.7T head-only MRI scanner they developed to make high-quality neuroimaging more widely available. Patients would be scanned while seated in the throne-like scanner, which sports a high-temperature superconducting magnet weighing 400 kg with a footprint of 6.2 sq m (67 sq ft). Researchers at the show demonstrated high-resolution scans of a lemon, for which they received a summa cum laude award. 

• Inexperienced Techs Linked to MRI Accidents: In another presentation at ISMRM 2024 (this one from the ISMRT section of the meeting), Japanese researchers found a link between MRI accidents and inexperienced MR technologists. They examined 26 accident reports over 10 years from 2012 to 2021 in which ferromagnetic objects were brought into the MRI suite, finding that half of the reports were related to technologists with less than one year of experience. The study highlights the importance of staff education and training in preventing accidents. 

• United’s MRI Advances: In industry news from ISMRM 2024, United Imaging followed up on the clearance of its uMR Jupiter 5T MRI scanner by announcing a pair of new MRI scanners. These include the uMR Sagitta 3T scanner with gradients rated at 240 mT/m and 200 T/m/s, and the uMR NX Frontier 3T system dedicated to brain research, with gradients rated at 650 mT/m and 600 T/m/s. The systems have not been submitted to the FDA and are not available for commercial or clinical use in the US. 

• GE Launches 3T Head MRI Scanner: Also at ISMRM 2024, GE HealthCare launched Signa Magnus, a 3T head-only MR scanner that’s focused on neuroscience research. Magnus sports GE’s HyperG gradient technology, with gradients rated at 300 mT/m and 750 T/m/s, and the scanner’s asymmetric gradient design supports superior diffusion performance, achieving extremely high B-value diffusion with short echo times (TEs). Magnus is pending 510(k) clearance; the first system has been installed at Brigham and Women’s Hospital for research use.

• Samsung to Buy Sonio: Samsung Medison has signed a deal to acquire French prenatal ultrasound developer Sonio. Sonio last year received FDA clearance for its Sonio Detect software, which guides sonographers through prenatal scans in real time to ensure exam completeness while also detecting potential fetal abnormalities. The companies said that on completion of the acquisition Sonio will remain an independent company, and its products and services will still be compatible with all ultrasound scanner vendors.    

• Boosting Breast Screening Compliance: UPenn researchers tried a variety of tools to boost mammography screening compliance, but one worked consistently: text messages to women about their appointments. In a study in JAMA Internal Medicine, researchers tested different referral and reminder strategies in 25k women, finding that women who got text messages had completion rates at three months at least two percentage points higher than those who didn’t. Combining bulk-order patient referrals and text reminders had the highest response rate compared to standard outreach (15.4% vs. 12.7%).   

• Stretching Prevents Tech Injuries: Preventing MSK injuries among allied health staff in interventional and echocardiography labs could be as simple as having them perform routine stretching exercises. In a study in JSCAI, Mayo Clinic researchers had 68 technologists and nursing staffers complete a 15-minute daily stretching routine for one year, after which participants reported a decrease in median Disability of Arm, Shoulder, and Hand (DASH) scores (5.2 to 2.6). Other recent research has documented the relationship between MSK problems and radiation protection gear.

• RT Hit with Fine in Credentials Suit: A radiologic technologist in Tennessee who the ARRT accused of misrepresenting her credentials was ordered to pay $125k in damages and attorney’s fees by a federal court. ARRT accused Lemesha Cannon of misrepresenting her status as a registered technologist and misusing ARRT trademarks when she was neither registered nor certified with the group. 

• CARPL Signs with Philips in Brazil: AI platform company CARPL.ai has signed a partnership with Philips in which its AI orchestration technology will be integrated with Philips’ Vue PACS enterprise imaging software in Brazil. The relationship will give Brazilian radiologists faster access to AI algorithms including for chest X-ray analysis, and will build on relationships CARPL already has in Brazil, such as with Hospital Israelita Albert Einstein (HIAE), Grupo Fleury, and Diagnósticos Da América (DASA).

• Siemens Extends Ohio Health System Pact: Siemens Healthineers has extended its ongoing relationship with University Hospitals of Ohio for another 10 years, with the new phase of the collaboration to focus on improving oncology, cardiovascular, and neurovascular care. Siemens and UH will also concentrate on theranostics and therapies for Alzheimer’s disease, and as part of the deal UH is purchasing Siemens PET/CT scanners as well as 1.5T and 3T MRI systems. Other research will emphasize MRI technologies such as Siemens’ 0.55T mid-field scanner.

• RapidAI Gets AI Perfusion Clearance: RapidAI has received 510(k) clearance for AngioFlow, an AI module that analyzes perfusion images of patients with suspected stroke in the interventional suite. AngioFlow produces qualitative perfusion maps that can help clinicians assess ischemic changes in brain regions with reduced blood flow. The AngioFlow nod builds on a series of clearances RapidAI received in 2023, such as for an algorithm to detect subdural hematoma (SDH). 

• Rad AI Bucks VC Funding Trend: VC funding of AI developers may be down, but that didn’t stop Rad AI from announcing a big $50M Series B funding round this week. The new round was led by Khosla Ventures and brings the company’s total capital raised to over $80M. The company said it would use the funding to accelerate product development and commercialization, such as of Rad AI Reporting, its AI-powered solution for radiology reporting workflow that was launched in September 2023. 

Faster MRI Scans with No Compromise

What if you could speed up your MRI scans with no compromise on image quality? In this video from RSNA 2023, we talked to SpinTech MRI CEO Karen Holzberger and Senior Clinical Sales Executive John Ciliberto about how their STAGE software is making this vision a reality.

Delivering Quality Patient Care

MEDRAD Stellant FLEX helps you meet your financial, clinical, and organizational goals, and it’s the only CT injection system currently 510(k)-cleared for use with contrast-enhanced mammography. Learn more about the benefits from Bayer.

The Resource Wire

• Automated Solutions for Patient and Provider Sharing: What’s the latest news from Clearpath? Find out how its fully automated patient and provider solutions for sharing medical records and images are making waves in healthcare.

• Catch Up with Live Fast 5: Missed any of Enlitic’s Live Fast 5 streams on LinkedIn? No worries! Catch all the action on their YouTube channel, including topics like data de-identification and anonymization, AI tools for monetization, efficient data routing, and much more. 

• Improving Patient Outcomes in Lung Cancer: Early detection is key to improving treatment pathways and patient outcomes in lung cancer. Learn more facts about lung cancer screening on this resource page from Riverain Technologies. 

• AI-Powered Informatics to Drive Healthcare Value: The DeepHealth OS is a cloud-native operating system that orchestrates all data to drive value across the healthcare enterprise. Learn more about how DeepHealth elevates the radiologist beyond radiology and across the entire care pathway.

• Radiology Automation Simplified: Validate, deploy, and discover radiology AI with CARPL.ai. Learn more about the company’s work in taking AI from bench to clinic by booking a demo today. 

• The Value of MRI Ownership: Are you wondering whether buying a new MRI scanner is the best use of scarce capital equipment funds? Check out Siemens Healthineers’ interactive ROI calculator to see the value of purchasing the MAGNETOM Free.Max. 

• How Imaging Partners of OC Eliminated CDs: Imaging Partners of Orange County was using CD-ROMs to share over 70,000 imaging exams they perform annually. Find out how PocketHealth helped them eliminate CDs and provide a better patient experience.

• Looking to Get Started with AI? Learn how Milton Keynes University Hospital adopted a successful AI strategy in this on-demand webinar from Blackford. You’ll discover how to evaluate and select the best AI solutions for your facility, tips for getting started with AI, and more. 

• Delivering on PACS Innovation: For over 20 years, Merge by Merative has delivered on PACS innovation. Merge PACS simplifies physicians’ reading activities, helps scale care delivery, and improves reading workflows across the enterprise. Learn more about Merge PACS – and keep your eyes peeled for an update later this year.

• The Journey to Cloud-Based PACS: Radiology is looking closely at cloud-based PACS as an alternative to on-premises digital image management. We discuss this emerging trend with key opinion leaders who have made the switch with Visage Imaging in The Imaging Wire Show.

• Strategic Planning for Radiology Groups: The pressures on radiology groups have never been more intense. In this May 21 webinar hosted by Medality, you’ll learn why strategic planning is important, how to start doing the planning process, and how to position yourself and your group for success. 

• Transforming Cardiovascular Ultrasound Analysis: Learn how cardiovascular ultrasound is on the verge of transformation with the latest version of Us2.ai’s new echo AI analysis software, featuring 45 automated echo parameters for streaming clinical workflow.

• AI Helps Residents Detect Fractures: Gleamer’s BoneView AI algorithm helped radiology residents detect fractures on adult and pediatric radiographs. Learn more about how it worked in this research summary. 

• Your In-Workflow Copilot Experience: PowerScribe Smart Impression from Nuance harnesses generative AI to automatically generate draft impressions and recommendations, saving you time and reducing workload-related stress. Learn how it can become your in-workflow copilot today. 

• Top-Tier Care at Rural Hospitals: Holzer Health System in Jackson, OH, treats local patients like family. In this video, learn how United Imaging equipped Holzer Health with its uMR 570 MRI scanner, helping them to offer top-tier care.