|
Imaging News from ESC 2024 | Mammo Wars Flare Up
September 5, 2024
|
|
|
|
|
Together with
|
|
|
|
“Beware of using ChatGPT for scholarly assistance. It just makes stuff up.”
|
|
Richard Duszak, MD, on ChatGPT’s hallucination effect when listing references.
|
|
|
|
The European Society of Cardiology annual meeting concluded on September 2 in London, with around 32k clinicians from 171 countries attending some 4.4k presentations. Organizers reported that attendance finally rebounded to pre-COVID numbers.
While much of ESC 2024 focused on treatments for cardiovascular disease, diagnosis with medical imaging still played a prominent role.
- Cardiac CT dominated many ESC sessions, and AI showed it is nearly as hot in cardiology as it is in radiology.
Major imaging-related ESC presentations included…
- A track on cardiac CT that underscored CT’s prognostic value:
- Myocardial revascularization patients who got FFR-CT had lower hazard ratios for MACE and all-cause mortality (HR=0.73 and 0.48).
- Incidental coronary artery anomalies appeared on 1.45% of CCTA scans for patients with suspected coronary artery disease.
- AI flexed its muscles in a machine learning track:
- AI of low-dose CT scans had an AUC of 0.95 for predicting pulmonary congestion, a sign of acute heart failure.
- Echocardiography AI identified HFpEF with higher AUC than clinical models (0.75 vs. 0.69).
- AI of transthoracic echo detected hypertrophic cardiomyopathy with AUC=0.85.
Another ESC hot topic was CT for calculating coronary artery calcium (CAC) scores, a possible predictor of heart disease. Sessions found …
- AI-generated volumetry of cardiac chambers based on CAC scans better predicted cardiovascular events than Agatston scores over 15 years of follow-up in an analysis of 5.8k patients from the MESA study.
- AI-CAC with CT was comparable to cardiac MRI read by humans for predicting atrial fibrillation (0.802 vs. 0.798) and stroke (0.762 vs. 0.751) over 15 years, which could give an edge to AI-CAC given its automated nature.
- An AI algorithm enabled opportunistic screening of CAC quantification from non-gated chest CT scans of 631 patients, finding high CAC scores in 13%. Many got statins, while 22 got additional imaging and 2 intervention.
- AI-generated CAC scores were also highlighted in a Polish study, detecting CAC on contrast CT at a rate comparable to humans on non-contrast CT (77% vs. 79%), possibly eliminating the need for additional non-contrast CT.
The Takeaway
This week’s ESC 2024 sessions demonstrate the vital role of imaging in diagnosing and treating cardiovascular disease. While radiologists may not control the patients, they can always apply knowledge of advances in other disciplines to their work.
|
|
|
Making Advanced Cardiac CT Accessible Everywhere
How do you bring cardiac CT to high-risk patients in rural areas? Watch this video from Siemens Healthineers to learn insights into setting up a cardiovascular CT program in a rural area – and how to overcome associated challenges.
|
|
AI-Driven Novice Heart Failure Screening
We hear a lot about AI’s potential to expand echocardiography to far more users and clinical settings, and a study using Us2.ai’s AI-automated echo analysis and reporting solution showed that echo AI might make novice-led heart failure screening possible.
|
|
- Age-Based Mammography Criticized: The wars over breast screening flared up this week with a JAMA Internal Medicine editorial criticizing the USPSTF’s recent decision to lower mammography’s screening age to 40. Researchers who in the past have advocated risk-based breast screening with genetic testing claim that starting screening at 40 “prevents few additional breast cancer deaths and adds substantial harms and costs.” They suggest that screening decisions for women younger than 50 should be based on breast cancer risk rather than age to optimize the benefits of screening while minimizing harms.
- False Positives and Future Screening: Another front in the debate over mammography are “harms” allegedly caused by breast screening. An Annals of Internal Medicine study suggested women who get false-positive mammograms may avoid future screening rounds. Of 1.1M women screened from 2005 to 2017, women were less likely to return after false positives that required additional imaging than true negatives, but the difference was small (75% vs. 76.9%). Screening rates declined more sharply if short-interval follow-up or biopsy was required (61% and 66.9%).
- Enlitic to Buy IT Pioneer Laitek: In a union of imaging IT firms, Enlitic has agreed to buy Laitek, an Illinois-based pioneer of PACS data migration and archiving solutions. Laitek was one of the first imaging IT firms, founded in 1980 by Fred Behlen, PhD, who still serves as president. The company provides contract-based data migration services and generated $6.8M in fiscal 2023 revenues from direct and OEM customers. On completion of the $5M acquisition, the combined companies should generate revenues of $8.3M-$9.6M in calendar 2024.
- Rezolut Offers Mammo AI to Patients: Imaging center firm Rezolut is offering AI analysis from Lunit to screening mammography patients as an optional service for which they can pay out-of-pocket. Called SecondReadAI, the service is available at two of Rezolut’s Southtowns Radiology locations outside Buffalo, New York, and will be rolled out to other sites in the company’s 40-center network. The partnership is a sign of growing acceptance of out-of-pocket payments for AI mammography interpretation, such as that offered by RadNet with its Enhanced Breast Cancer Detection service.
- ChatGPT Struggles with Medical Images: A version of ChatGPT designed for image analysis struggled with radiology images in a new study in Radiology. ChatGPT-4 Vision is the first version of the large language model that can interpret both text and images, but in tests on 377 questions from an ACR exam for residents the LLM correctly answered a higher percentage of text-only questions (82%) compared to questions with images (48%). ChatGPT-4 Vision performed best on image-based questions in the chest (69%) and genitourinary (67%) specialties.
- Harrison Launches Radiology Vision Model: Could better performance occur with vision language models developed specifically for radiology? Harrison.ai is giving it a shot with Harrison.rad.1, a radiology-specific vision language model that it’s providing to industry partners, healthcare professionals, and regulators. The company said the dialogue-based model was trained on real-world imaging data and can detect and localize radiology findings, generate reports, and provide longitudinal reasoning based on clinical history and patient context. Harrison is the parent company of Annalise.ai.
- RSNA Data Breach: The RSNA has notified several state authorities of data breaches that may have affected people in those states. RSNA said that a cyberattack in June 2024 resulted in “unauthorized access” to its network, with potential exposure of “a limited number of 1099 forms.” The exposure exclusively affected about 200 contractors to whom RSNA has sent payments for services rendered, and not the group’s entire membership.
- AI for ARIA Approved in Europe: icometrix has received European approval for icobrain aria, the company’s AI algorithm for detecting amyloid-related imaging abnormalities (ARIA) on MRI scans. ARIAs are abnormal findings that can occur in patients with Alzheimer’s disease who are taking drugs like Leqembi that target amyloid; clinical guidelines advise MRI scans for monitoring, but they can be challenging for radiologists to detect and assess. A recent research study found that icobrain aria was effective in detecting different ARIA types.
- Cortechs Clearance for AI Update: The FDA’s been busy clearing ARIA-related algorithms. Cortechs.ai announced FDA clearance of NeuroQuant 5.0, a new version of its software that includes enhancements to help radiologists perform better segmentation of ARIAs on MRI scans. The capabilities come from advanced segmentation and quantification functionality built into the new version for various MRI sequences; in addition to ARIAs, the new version addresses TBI and cerebral amyloid angiopathy.
- Segmed Raises $10M: Imaging data aggregation firm Segmed raised $10.4M in a Series A funding round (total funding now >$20M) led by a combination of new and current investors. The company said it would use the funds to expand in the healthcare AI space, as well as to bring its medical imaging technology into the biopharmaceutical and life sciences industries. Segmed’s Openda platform gives researchers and software developers access to de-identified medical imaging data and includes tools to help users unlock insights from healthcare datasets.
- Blackford Signs Coreline to Platform: Blackford has added Coreline Soft’s AI solutions for thoracic CT to its Blackford Platform. Coreline’s AVIEW family of AI algorithms will be available to Platform customers, including AVIEW LCS for identifying and assessing pulmonary nodules; AVIEW COPD for detecting and classifying emphysema and chronic obstructive pulmonary disease; and AVIEW CAC for scoring coronary artery calcium deposits. All three solutions are available in Coreline’s AVIEW LCS Plus package for identifying all three most common lung diseases.
- Samsung Finalizes Sonio Acquisition: Samsung has completed its $95M acquisition of French prenatal ultrasound AI software developer Sonio, a move first announced in May. Sonio Detect guides sonographers through prenatal scans in real time to ensure exam completeness while also supporting real-time reporting of potential fetal abnormalities; the software received FDA clearance in 2023, with an updated version cleared in April. The purchase highlights the trend of ultrasound makers buying echo AI firms, including GE HealthCare (Caption Health), Exo (Medo AI), and Philips (DiA Imaging Analysis).
- NIH Stunner on Havana Syndrome: The NIH has canceled its investigation into the mysterious Havana syndrome, a set of neurological ailments suffered by U.S. embassy personnel in Cuba. A CNN report said the NIH stopped the research after learning that the CIA “coerced” victims into participating in research studies as a requirement for getting healthcare. NIH researchers in a March 2024 study in JAMA claimed there was no evidence on MRI scans of neurological damage in victims, contradicting older studies that found evidence of brain changes.
|
|
AI Supports Diagnosis of Cognitive Disorders
How can AI support the early diagnosis of cognitive disorders in elderly people? Read this article from DeepHealth on how AI-powered software with MRI can be used with both neuropsychological tests and clinical information to improve early diagnosis of cognitive impairments.
|
|
Preparing for the Future of Enterprise Imaging
Unlock medical imaging’s potential with the power of cloud technology. Click here to discover how Optum’s solutions can streamline your operations, enhance data security, and improve patient outcomes. Embrace medical imaging’s future and see the difference the cloud can make.
|
|
Stop Shipping Discs!
By pivoting to a 100% digital fulfillment model for patient images and records, you can improve their experience while significantly reducing labor and shipping costs. Find out how on this page from Clearpath.
|
|
- Data Standardization and Patient Care: Standardization of patient data is a crucial aspect of healthcare. It helps make practices more efficient, but it also benefits patients by improving outcomes and patient satisfaction. Learn how data standardization can help improve your practice in this article from Enlitic.
- 2 Questions about AI for Radiology Leaders: Are today’s radiology AI solutions solving the right problems? And are there other solutions available for AI of brain MRI? Read this article from SpinTech MRI to learn how its STAGE solution can optimize MRI utilization.
- Unify Imaging Workflow: Radiology group Qscan faced a fragmented ecosystem after the rapid acquisition of eight companies, each with their own reporting systems and toolsets. Learn how InteleOrchestrator from Intelerad helps them unify their imaging workflow.
- AI for Lung Cancer Diagnosis and Screening: Check out this comprehensive new eBook from Calantic by Bayer on the role of AI in lung cancer diagnosis and screening. It explores AI’s potential role in improving lung cancer screening strategies, identifying high-risk individuals, and enhancing diagnostic accuracy. Download it today.
- White Paper on Structural Heart Imaging: How can AI help improve care for patients with structural heart issues? Learn more about new enhancements for procedures like TAVR and TMVR, as well as LAA closure, in this downloadable white paper from TeraRecon.
- Learn the Basics of CT Colonography: Medicare will soon be paying for CT colonography exams for colon cancer screening. Get up to speed today with Medality’s comprehensive CT colonography course, led by experts Judy Yee, MD, Kevin Chang, MD.
- Celebrating Happy Customers: There’s nothing better than happy customers. Find out what radiology personnel at Western New York MRI in Buffalo had to say about their new uMR 680 scanner from United Imaging.
- Set a New Standard of Care: Give your patients clarity and peace of mind with Prenuvo’s advanced whole-body MRI scans. Prenuvo’s team works closely with you to help pinpoint potential problems, diagnose faster, and deepen your patient relationships. Learn more today.
- Next-Generation Radiology Reporting: Join the conversation in this webinar recording and hear from PACS administrator Griff R. Van Dusen of Memorial Health System how Nuance PowerScribe One’s next-generation reporting experience helps streamline workflow and improve report quality so radiologists can get more done in less time.
- Automated Weight-Bearing Foot Measurements: Learn how Gleamer’s BoneMetrics AI software was able to provide automated measurements on weight-bearing radiographs with high levels of accuracy in a paper published in Skeletal Radiology.
- AI and Matters of the Heart: As clinicians face increasing pressure from all sides, can AI provide some much-needed breathing space? In this article from Blackford, learn more about how AI is assisting with cardiac imaging with modalities ranging from CT to echocardiography.
- Drivers of AI Usage in Radiology: Radiologists are being asked to read more, read faster, and with a higher degree of accuracy as imaging data grows in volume and complexity. In this video, listen to Bernardo Bizzo, MD, and Riverain CEO Steve Worrell explain how this is driving AI usage.
- Enhancing Patient Experience and Streamlining Operations: Jefferson County Health Center transformed their image exchange by eliminating the need for CDs, saving over 1,000 staff hours annually. This shift improved radiology patient satisfaction scores by 7.6%. Learn how JCHC enhanced patient care and operational efficiency with PocketHealth.
- 5 Things to Know about Cloud PACS: The cloud doesn’t need to be hard. Visage 7 from Visage Imaging delivers simplicity with proven experience. Learn about the five things you need to know in this video.
- Introducing Merge Consulting Services: How can your radiology practice manage change and expectations in healthcare with workforce flexibility? Merge by Merative is launching Merge Consulting Services to help you face the challenges of sourcing expertise and staff. Learn more about it today.
|
|
|
|
|
