“47 times over the last year you ordered a CT scan with an indication that just said ‘pain'”
Some feedback for surgery during Dr. Glaucomflecken’s latest faculty meeting.
Imaging Wire Sponsors
Arterys | Bayer Radiology | Canon Medical Systems | GE Healthcare
Healthcare Administrative Partners | Hitachi Healthcare Americas
Novarad | Nuance | Riverain Technologies | Siemens Healthineers
United Imaging | Zebra Medical Vision
The Imaging Wire
Opportunistic Cardiac Screening
A new study out of Stanford detailed a coronary artery calcification (CAC) scoring system that could evolve into an “opportunistic screening” pathway leveraging existing chest CT scans.
- Background – Over 20m chest CTs are performed in the U.S. annually and that number is poised to grow even higher with the USPSTF’s expanded LDCT screening guidelines. Each of these scans contains insights into patients’ cardiac health. However, an AI model like this would be required to extract cardiac data from the majority of CT scans (CAC not visible to humans w/ non-gated CTs) or efficiently interpret them (there’s far too many images).
- The CAC Models – The Stanford team first developed a deep learning model to produce CAC scores using gated coronary CT exams (training: 697, validation: 78), and then developed a second model with routine non-gated chest CTs (training: 314, validation: 45).
- The Results – The gated coronary CT model achieved near perfect agreement with manually-generated CAC scores, and produced these scores in just 3.5 seconds (vs. 261 seconds avg. from three technologists). The second non-gated model accurately identified patients with any CAC (sensitivity: 80–100%; PPV: 87–100%) and patients with CAC ≥ 100 (sensitivity: 71–94%; PPV: 88–100%).
- An Opportunistic Use Case – A model like this could help expand CAC screening to far more patients and lead to more early cardiac interventions, using the images we already have.
- A Population Health AI Milestone – This study adds to the growing field of evidence supporting imaging AI’s role in population health (here’s another from last week). And because this study comes from a particularly admired and social-savvy group, it could prove to be a key milestone for population health AI awareness.
Take the Canon AiCE Challenge
Take the AiCE challenge and see why half the radiologists in a recent study “had difficulty differentiating” images from Canon Medical Systems’ Vantage Orian 1.5T MR using its AiCE reconstruction technology compared to standard 3T MRI images.
Bayer Goes Beyond Compliance
This Bayer case study details how radiation benchmarking programs can help push CT dose exposure reduction initiatives from achieving compliance to driving quality.
- AI Shortcuts: A newly-published U of Washington study detailed a series of AI models that rely on shortcuts to detect COVID-19 in CXR scans rather than actual pathology (patient position, age, etc.). We already detailed this study last fall when it was a pre-print, but the new version also revealed that models built with more carefully constructed datasets will generalize better to external hospitals.
- CompuGroup’s PACS Acquisition: Major global healthcare IT company, CompuGroup Medical, expanded into the radiology arena with its acquisition of German PACS and healthcare content management company, VISUS Health IT (€18.5m 2020 revenue, used at 1.5k sites). CompuGroup will combine VISUS Health IT with its own medical information systems business, expanding into VISUS’s existing accounts and bringing VISUS’s PACS/HCM solutions into its own accounts.
- Cincinnati Children’s Enterprise POCUS: A SIIM 2021 presentation detailed how Cincinnati Children’s Hospital implemented point of care ultrasound into its enterprise imaging system (fast forward to 0:30), allowing $2.5m in new POCUS service billings. To do this, Cincinnati Children’s: 1) Created a POCUS implementation team; 2) Evaluated its POCUS devices for EI readiness; 3) Credentialed its clinicians for POCUS use; 4) Established its POCUS billing procedures; 5) Built POCUS into its EMR; 6) Trained its users; and 7) Started rolling-out POCUS EI across its departments.
- Lantheus PCa PET Paradigm Shift: Lantheus Medical Imaging announced the FDA approval of its PYLARIFY (F-18 DCFPyL) PET agent for prostate cancer, calling it the “first and only” commercially available PSMA-targeted PET agent. PYLARIFY is highlighted by its ability to detect prostate cancer’s spread to distant organs (a challenge with other agents), allowing earlier and more accurate treatment planning.
- Staying Independent: Healthcare Administrative Partners just shared a helpful set of guidelines that radiology practices can follow to stay private amid the ongoing consolidation trend. HAP encourages staying-private practices to: 1) Identify / address their opportunities and weaknesses; 2) Leverage peer and vendor partnerships to improve efficiency / scale; 3) Deepen client relationships by delivering more value; and 4) Continue to provide great service to referring physicians and patients.
- Kailo & Medo Integrate: Kailo Medical will integrate Medo AI’s Medo-Thyroid ultrasound workflow solution (FDA-approved, analyzes/categorizes thyroid ultrasound) into its SonoReview structured reporting software, which they say will improve thyroid ultrasound efficiency and accuracy.
- YNHH’s Case for #DitchingtheDisk: Last week’s SIIM 2021 meeting included an impressive case study from Yale New Haven Health (fast forward to 1:13), detailing how its transition to a Nuance electronic imaging sharing system eliminated $650k in annual CD/DVD costs. After spending over $550k to produce 142k CD/DVDs in 2019 ($3.95 each, not including film library labor and shipping), YNHH launched its Zero CD Initiative (IT, workflows, policies, training, outreach), resulting in all 165k of its 2020 studies being shared through its image exchange platform.
- CINA CHEST Approved: Avicenna.AI announced the FDA and CE approval of its CINA CHEST AI solution, which analyzes CT angiography scans to detect and triage pulmonary embolism and aortic dissection (PE & AD). CINA CHEST joins CINA HEAD (detects/triages stroke and neurovascular emergencies) in Avicenna.AI’s growing emergency triage portfolio.
- COVID LUS’ Low Barrier: Clinicians with various levels of experience can adequately perform lung ultrasound COVID assessments, although training helps. That’s from a German study that had ten observers with different subspecialty backgrounds assess 100 lung ultrasound cine-loops from hospitalized COVID patients (4x each, 400 loops overall), revealing moderate-to-substantial agreement across the observers and achieving the highest agreements with the most distinct findings (e.g. low & high LUS score).
- Radiology’s PA Exposure: A new JAMA study (n = 6.47m beneficiaries) revealed that if Medicare Part B adopted similar prior authorization rules as private payers, radiology would be among the most exposed. Roughly 91% of radiologists’ Medicare Part B services would have required PAs (3rd after oncologists’ 97% and cardiologists’ 93%), although that would be partially justified by radiology’s spot as the largest source of non-drug spending (16% of total).
- Xenon Lung MRI: Researchers from Ontario’s Western University detailed a Xenon-based lung MRI technique that they say could “revolutionize” detection and evaluation of a range of lung diseases (and potentially other organs). The technique involves patients inhaling the hyperpolarized xenon-129 agent, resulting in MRI scans that detail structural and functional issues “exponentially better” than standard MRIs.
- Visionairy’s CXR AI CE: Visionairy Health secured CE Marking for its X1 Chest X-ray triage and prioritization tool (Class I, self-certified), which flags 15 different CXR findings including TB, pneumothorax, and lung nodules. Visionairy Health is an earlier stage startup than most regulatory-approved AI players (~$650k in funding, 5 employees on LinkedIn), but views this as the first step in a wider global and product expansion.
- Cancer Imaging’s High Stakes: A new JACR study revealed that oncology is the most common source of diagnostic radiology malpractice cases. The review of 2008-2017 U.S. malpractice claims, found that radiology was responsible for 12.8% of all diagnostic-related malpractice claims (1,756 of 13,695), while oncology accounted for 44% of those radiology diagnostic cases (772 of 1,756), largely due to imaging misinterpretation (623 of 772).
- St. Joseph Mercy Follows-Up: A SIIM 2021 presentation from St. Joseph Mercy Health detailed how the Detroit-based system significantly improved its incidental lung nodule follow-up rates. After finding that just 36% of its incidental lung nodules were tracked, St. Joseph Mercy developed a follow-up system that combined new team procedures (radiologists recommend follow-ups, nurses track them), new communication structures (patient records flagged, ongoing follow-ups recorded), and a Nuance NLP-based system to guide this process. St. Joseph Mercy Health’s follow-up closure rate is now nearly 99%.
Hitachi’s CVIS Performance & Outcomes
Learn how leveraging the right cardiology image and reporting platform drives performance and outcomes in this Hitachi white paper.
The Resource Wire
- Can Cloud Security keep healthcare data secure? Read this whitepaper to find out how Arterys utilizes the cloud to keep this data private and protected.
- Check out this blog from the NIH’s Dr. Francis Collins, where he discusses how United Imaging’s “groundbreaking” Explorer total-body PET/CT “opens a new window into human biology.”
- The Saint Joseph Mercy Health System doubled its follow-up recommendation identification/tracking when it adopted Nuance PowerScribe Follow-up Manager, achieving ROI within the first year. Find out how in this Nuance case study.
- See why AI-powered cardiovascular screening’s detection and prevention benefits make it the new frontier of population health.
- Novarad’s COVID-19 AI Diagnostic Assistant was named the ‘Best New Radiology Solution’ for its ability to quickly and accurately diagnose COVID-19 patients.
- Know how your practice measures up? In this post, Healthcare Administrative Partners details the key benchmarking quality metrics and how they can help radiology practices improve.
- See how physicians are leveraging GE Healthcare’s Edison Open AI Orchestrator platform and Icobrain’s AI-enabled applications to monitor and treat MS and Traumatic Brain Injury.
- Make quick and well-informed decisions by having a clear overview of your clinical and operational performance data with Siemens Healthineers’ teamplay.
- Watch Jared Christensen, MD, MBA explain how Duke University Health uses Riverain Technology’s ClearRead CT Vessel Suppress and ClearRead CT Detect in its daily practice.