|
AI Predicts Cancer Before It’s Detected | SABCS Boycott?
October 7, 2024
|
|
|
|
|
Together with
|
|
|
|
“An AI system that indicates the woman’s individual risk for breast cancer based solely on mammograms could provide a streamlined, more efficient approach to risk-based screening decisions.”
|
|
Gjesvik J et al, on findings from a new study using AI to predict breast cancer risk.
|
|
|
Patient-centric image exchange is empowering patients and improving engagement with their healthcare providers by giving them easier access to their own images and medical data. We talked about the benefits of patient-centric image exchange with Rishi Nayyar, co-founder and CEO of PocketHealth, in this episode of the Imaging Wire Show.
|
|
|
|
A new study highlights the predictive power of AI for mammography screening – before cancers are even detected. Researchers in a study JAMA Network Open found that risk scores generated by Lunit’s Insight MMG algorithm predicted which women would develop breast cancer – years before radiologists found it on mammograms.
Mammography image analysis has always been one of the most promising use cases for AI – even dating back to the days of computer-aided detection in the early 2000s.
- Most mammography AI developers have focused on helping radiologists identify suspicious lesions on mammograms, or triage low-risk studies so they don’t require extra review.
But a funny thing has happened during clinical use of these algorithms – radiologists found that AI-generated risk scores appeared to predict future breast cancers before they could be seen on mammograms.
- Insight MMG marks areas of concern and generates a risk score of 0-100 for the presence of breast cancer (higher numbers are worse).
Researchers decided to investigate the risk scores’ predictive power by applying Insight MMG to screening mammography exams acquired in the BreastScreen Norway program over three biennial rounds of screening from 2004 to 2018.
- They then correlated AI risk scores to clinical outcomes in exams for 116k women for up to six years after the initial screening round.
Major findings of the study included …
- AI risk scores were higher for women who later developed cancer, 4-6 years before the cancer was detected.
- The difference in risk scores increased over three screening rounds, from 21 points in the first round to 79 points in the third round.
- Risk scores had very high accuracy by the third round (AUC=0.93).
- AI scores were more accurate than existing risk tools like the Tyrer-Cuzick model.
How could AI risk scores be used in clinical practice?
- Women without detectable cancer but with high scores could be directed to shorter screening intervals or screening with supplemental modalities like ultrasound or MRI.
The Takeaway
It’s hard to overstate the significance of the new results. While AI for direct mammography image interpretation still seems to be having trouble catching on (just like CAD did), risk prediction is a use case that could direct more effective breast screening. The study is also a major coup for Lunit, continuing a string of impressive clinical results with the company’s technology.
|
|
|
Request an RSNA Meeting with TeraRecon
RSNA 2024 will be here before you know it. Come explore TeraRecon’s latest updates and find out why the company is an award-winning solution provider for AI-empowered radiology, oncology, cardiology, neurology, and vascular surgery.
|
|
Enterprise Imaging for Radiologist Well-Being
With the ongoing shortage of radiologists and ever-increasing imaging volume, burnout has been a persistent problem facing radiology. Join this October 9 webinar hosted by AGFA HealthCare to hear how radiology is turning to enterprise imaging and AI to improve its workflow.
|
|
Connect with United Imaging at RSNA 2024
United Imaging will be celebrating the theme of Building Intelligent Connections at RSNA 2024. Come visit the company at booth #1929 to learn about their imaging solutions and how they connect to United’s mission of Equal Healthcare for All.
|
|
- Breast Surgeon to Boycott SABCS: Breast cancer surgeon Laura Esserman, MD, is getting attention for her boycott of the upcoming San Antonio Breast Cancer Symposium over abortion policies in Texas. Esserman is best known for her research on risk-based strategies for breast screening but has also been a proponent of women’s rights; she began boycotting SABCS after Texas passed an abortion ban in 2021. While breast cancer specialists have expressed reservations about SABCS’ location in the past, Esserman is the highest-profile researcher to boycott the December 10-13 conference.
- DBT vs. 2D Mammography Redux: A new meta-analysis comparing breast screening with DBT to conventional 2D mammography produced conflicting results. Researchers in European Radiology analyzed 10 prospective large-scale studies with a total of 205k DBT-screened patients and 306k 2D-screened patients. There was no difference in the interval cancer rate (cancers between screening rounds), which could have been caused by biased study design that drew patients from different time periods or locations. DBT’s cancer detection rate on routine screening was 42% higher but there was no difference in recall rates.
- CEM for Women with Extremely Dense Breasts: Contrast-enhanced mammography could be an alternative for screening women with extremely dense breast tissue. Writing in a study in Radiology, researchers compared low-energy CEM (equivalent to 2D digital mammography) to high-energy CEM for screening 609 women with extremely dense tissue. High-energy imaging had much higher sensitivity (89% vs. 28%) but lower specificity (89% vs. 96%), although specificity improved with follow-up screening rounds. High-energy CEM’s recall rate was high (19%), but it could be an alternative to MRI, which is expensive and less accessible.
- Implant Risk During MRI Scans: Some patients with active implanted devices getting MRI scans could be at risk of complications due to missing implant safety information. In a study in Current Problems in Diagnostic Radiology, researchers at Mayo Clinic in Phoenix found there was data missing about either the patient or the implant for 16% of 749 patient records they reviewed. Some 78% of these cases were caused by missing implant vendor manuals, leading authors to recommend MRI sites keep a list of links to vendor manual repositories.
- Siemens Installs First U.S. Pro.Pulse CT: Siemens Healthineers achieved a milestone in its drive to make dual-source CT available to a wider range of facilities with the first U.S. installation of its Somatom Pro.Pulse scanner, at Advanced Imaging of Montana. First introduced at RSNA 2023, Pro.Pulse takes high-end dual-source architecture – with two X-ray tubes and detector arrays – and bundles it into a more affordable package, while also adding in embedded AI and workflow assistance. Key to the design is an air-cooled rather than water-cooled gantry.
- Boosting Outpatient CT Volume: Imaging facilities worldwide are under pressure from rising exam volume. The University of Rochester Medical Center responded with several quality improvement initiatives that boosted outpatient CT volume and reduced scheduling times from six weeks to one week. The radiology department temporarily double-booked CT appointments, reduced scanning slot duration (to 15 minutes from 20), and standardized contrast protocols. Weekly exam volume jumped 19% (722 to 860), generating an extra $1.6M in revenue, and the scheduling interval fell to three days after they acquired a new CT scanner.
- Bunkerhill’s Gated CAC Clearance: Bunkerhill Health expanded its CAC-scoring AI portfolio, landing FDA clearance for a new solution that analyzes gated and non-gated CT exams to automatically identify patients with coronary artery calcium. Bunkerhill Health already has a solid list of customers using its non-gated CAC scoring solution, and the new clearance will allow health systems to analyze all gated and non-gated chest CTs that they perform, leading to earlier CVD detection and management. Separately, Bunkerhill landed an installation deal with Stamford Health’s Heart & Vascular Institute.
- Researchers Update CT Screening Paper: Researchers from China have published corrections to a paper on opportunistic low-dose CT lung cancer screening published in JAMA Network Open at the end of 2023. The authors said they received a number of comments on the paper’s methodology that prompted them to re-examine their data and revise their figures downward on screening’s life-saving impact. CT screening still has a positive impact, but the effect’s size is lower for lung cancer mortality (34% vs. 49%) and all-cause mortality (28% vs. 46%).
- Fracture AI Improves Over Time: An AI algorithm for detecting fractures on X-rays in the emergency department showed steady performance improvement over multiple versions. In a study from Spain, researchers tracked the performance of three successive versions of Gleamer’s BoneView algorithm in 2.7k adult and pediatric bone trauma X-rays. The algorithm’s positive predictive value improved 24% (from 57% to 70%) while specificity increased 4.8% (86% to 90%). Negative predictive value and sensitivity remained high across all three versions (98% and 94%, respectively).
- CDC Eyes AI for TB: Tuberculosis has emerged as a promising use case for AI of chest X-rays – and the CDC is testing the combination as part of the admissions process for immigrants and refugees. In a new study in PLOS Digital Health, CDC researchers developed three deep learning algorithms to identify abnormalities on chest X-rays, finding that they turned in good performance for detecting general abnormalities (AUC range 0.89-0.92) and findings suggestive of TB (AUC range 0.94-0.99). TB AI could be used as a QC tool.
- GE’s MIM Lands Theranostics Nod: GE HealthCare’s MIM Software business received FDA 510(k) clearance for Monte Carlo dosimetry to be used for planning theranostics treatments. The clearance supports the use of absorbed dose calculations for therapeutic radionuclides using Monte Carlo dosimetry with the Dose Planning Method, and will be available on MIM’s SurePlan MRT solution for automating and standardizing theranostics dosimetry. Theranostics is gaining momentum as a more precise method for hard-to-treat tumors and other diseases.
- Positrigo Brain PET Gets CE Mark: Hot on the heels of its 510(k) clearance in July, Swiss nuclear medicine developer Positrigo has received the CE Mark for its NeuroLF dedicated brain PET scanner. NeuroLF is targeted at neurological conditions like Alzheimer’s disease, brain tumors, and epilepsy, and the company sees the system as a more economical alternative to whole-body PET/CT scanners for brain imaging. Positrigo hopes to capitalize on the growing need for amyloid PET scans to monitor progression of patients being treated with new Alzheimer’s drugs.
- AI Alliance Targets Cancer Research: An array of healthcare industry heavyweights are joining forces to leverage the power of AI to fight cancer. The Cancer AI Alliance unites four NCI-designated cancer centers – led by Fred Hutch Cancer Center – with technology titans AWS, NVIDIA, Microsoft, and Deloitte to collaborate in using AI and massive computational power to solve research challenges and spur new cancer discoveries, such as by setting up a federated AI learning network to share data. CAIA is being fueled by $40M in seed money.
- FDA Issues Final X-Ray Rules: The FDA issued final radiation control rules for manufacturers of diagnostic X-ray equipment. The new guidance is an update of rules first issued in 1989, and while they don’t seem to be a massive rewrite of the agency’s guidelines it could be a good idea for radiology vendors to familiarize themselves with the update. The FDA first proposed the new rules in 2018.
|
|
An Integrated Approach to Radiology AI
AI automates what radiologists can’t stand, surfaces what radiologists can’t see, and identifies what radiologists can’t miss. But only if it’s implemented in the way radiologists work. See how Nuance helps radiologists achieve these results through a single, streamlined, end-to-end AI experience.
|
|
Unprecedented Insights Made Possible with AI
With the largest normative dataset of whole-body imaging in the world, Prenuvo’s AI researchers partner with the best academic minds to understand – like never before – what “normal” aging means. Learn about their work today.
|
|
How McFarland Clinic Reduced Downtime
McFarland Clinic Health Ventures recently experienced a downtime occurrence when an HL7 interface went down. But thanks to a suite of medical imaging solutions from Merge by Merative, McFarland’s downtime lasted all of 15 minutes. Find out how they did it in this case study.
|
|
- The Clinical Value of Soft-Tissue Chest X-Ray: Soft-tissue techniques can improve the visibility and accuracy of chest X-ray. Learn about two important soft-tissue methods – bone suppression and dual-energy subtraction – in this white paper from Riverain Technologies.
- Improving Your Mammography Workflow: Want to learn how AI can relieve pressure in your mammography department? Watch this on-demand webinar hosted by Blackford featuring a panel discussing the importance of assessing breast density, image quality reporting, and prioritizing studies.
- Transforming Patient Care with Mach7’s eUnity Diagnostic Viewer: University of Michigan Health-West uses Mach7’s eUnity Diagnostic Viewer to drive patient care and provider satisfaction. Learn more about how one of the nation’s most-wired hospitals puts patients first through adoption of innovative imaging technology.
- Simplifying Complex Image Exchange Workflows: Guadalupe Regional Medical Center (GRMC) and Methodist Hospital implemented a PocketHealth Community Gateway that saved over 1,700 staff hours. Read how they streamlined bidirectional image exchange, created operational efficiencies, and improved continuity of care for patients.
- Unlocking the Full Potential of Medical Imaging Data: Many healthcare organizations aren’t realizing the full potential of their vast stores of imaging data. On this page from Enlitic, learn how you can turn your data from an underutilized asset into a powerful driver of clinical, operational, and strategic value.
- Leave No Breast Cancer Patient Behind: Breast cancer is the second leading cause of cancer death among women, but catching it early greatly improves the survival rate. Leave no patient behind with the help of Intelerad’s InteleScreen and IntelePACS for breast imaging organizations.
- AI for Limb Fractures on X-Ray: AI can recognize limb fractures on X-rays and reduce interpretation discrepancies between radiology and emergency departments. Learn how Gleamer’s BoneView AI algorithm performed in this new research study.
- Clarity, Speed, and Confidence for MRI Efficiency: Radiologists have used a variety of methods to improve efficiency, but many of these methods come with drawbacks. Find out in this article how SpinTech MRI takes on the challenge of MRI efficiency with its STAGE software.
- Get the 2024 Radiology Practice Development Report: Medality surveyed more than 3,300 radiologists and discovered the most critical training gaps and growth opportunities in radiology for its 2024 Radiology Practice Development Report. Download your complimentary report today.
- AI and Cancer Screening: Cancer screening saves lives, but right now screening is limited to a few cancer types. That could change with AI, which opens new possibilities for earlier disease detection. Learn more in this article by DeepHealth clinical AI leader Greg Sorensen, MD.
- Digital Tools for Heart Failure: Clinicians have a growing array of digital tools for assessing patients with suspected heart failure. A new review article in Lancet Digital Health takes stock of some of the options, including echo AI tools like those from Us2.ai.
- Focus on Mammography Workflow: Mammography workflow is key to providing high-quality breast imaging services. In this Imaging Wire Show, we talked with Christie Devine of Siemens Healthineers about how recent advances in mammography workflow are leading to more effective mammography technologists — and happier patients.
- Catch What You’ve Missed on Unboxing AI: Missed an episode of Unboxing AI, CARPL’s video series on AI in radiology? Check out all the past episodes on their YouTube channel.
- Transform Healthcare with the Cloud: Discover how Optum’s cloud-based medical imaging solutions can slash costs and streamline radiology operations. Visit their website today to unlock a wealth of healthcare insights and resources tailored just for professionals like you.
- Imaging Workflows that Actually Work: Not a fan of medical image exchange on discs? Then check out Clearpath and find out how it’s removing obstacles to better radiology workflow. Request a demo today.
|
|
|
|
|
