CT’s Case | MDR Live | Disrupting NHS

“In this profession, nothing is certain except death, taxes, and reimbursement cuts.”

Andrew Wilmot, MD and Saurabh Jha, MBBS on why protecting reimbursements is a poor reason to maintain radiologists’ contrast reaction monitoring duties.

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

CT Appendicitis Scoring’s Low Scores

Guidelines may recommend using clinical scoring systems to triage patients with suspected appendicitis before ordering a CT scan, but a new study out of South Korea finds that these scoring systems are far from effective.

  • The Study – The retrospective study triaged 2,888 patients with suspected appendicitis (1,088 with & 1,800 without) using five diffident scoring systems, measuring their accuracy and effect on CT volumes.
  • The Results – The scoring systems were very effective at reducing CT exams (55.6% to 71.1% reductions), although they would have cancelled hundreds of appendicitis-positive patients’ CTs (sensitivity: 48.7% to 81.2%) and still allowed hundreds of unnecessary scans (specificity: 79% to 97.8%).
  • The CT Advantage – Radiation exposure concerns aside, CT’s 0.98 AUC with this patient group was way higher than the various scoring systems (0.64 – 0.75 AUCs).
  • The Takeaway – Just about every appendicitis CT study that we cover focuses on over-imaging, so it makes sense to consider new/improved triage methods, but this study suggests that it would be more effective to focus on “lowering the CT radiation dose rather than avoiding the use of CT itself.”

Advancing CMR with GE AIR Recon DL

See how deep learning image reconstruction technology is expanding how radiologists and cardiologists use cardiac MRI in this GE Healthcare report.

– Sponsored.

MDR Goes Live

Europe’s Medical Device Regulations (MDR) went live last week, increasing regulatory requirements across nearly all medical devices and software categories.

  • High Class AI – MDR now requires healthcare AI products to attain higher risk classifications (IIa, IIb, or III…. no longer class I), eliminating self-certifications and asking AI vendors to provide far more evidence of their efficacy.
  • MDR AI Timeline – Existing Class I AI products have three years to meet the new requirements, but the MDR rules apply to all new products and notable updates to current products.
  • MDR’s AI Impact – Considering that the AI for Radiology database only lists four brands with MDR-compliant products, many AI players will have a lot of regulatory work to do.
  • AI Reactions – Europe’s established AI players seemed to welcome this move, calling it a starting point for a “professionalized and mature” European AI marketplace.
  • Hardware MDR – The MDR will require most diagnostic imaging devices to meet Class IIa specifications, although there’s a lot less online chatter about MDR’s impact on hardware devices. That’s probably because most hardware OEMs were a lot better prepared for this change.

Summa and Nuance’s Follow-up Rebound

Check out how Nuance’s mPower Clinical Analytics solution helped Summa Health improve its incidental lung nodule follow-up rates by nearly 8x.

– Sponsored.

The Wire

  • Contrast Coverage Contention: The debate over whether radiologists should monitor contrast reactions at outpatient imaging centers resurfaced with a fierce rebuttal from the “against” authors. Here’s their counter argument: 1) It can hurt smaller practices; 2) Non-physicians commonly handle anaphylaxis; 3) Contrast monitoring is a bad reason to train more radiologists; 4) Monitoring keeps rads from practicing medicine “like real doctors;” 5) Fear over reimbursements and perception should not influence contrast monitoring policies.
  • Ultrasound AI for ITFA: A JAMA study detailed an ultrasound AI model that accurately quantified interstitial fibrosis and tubular atrophy (ITFA – indicates kidney decline), suggesting that a similar method could become the first noninvasive ITFA exam. The algorithm segmented ultrasound images from a 352-patient independent test set with 90% accuracy, performing consistently across different patient scenarios.
  • SIIM 2021: SIIM just wrapped what hopefully will be one of radiology’s last all-digital annual meetings, and in many ways SIIM 2021 showed how a great digital event can be done (or how future hybrid events should be done). They used a singular platform for all meetings and events, filled it with high-value content (presentations, panels, studies), held most segments live, and did their best to keep it collaborative (real-time chat/questions, attendee hackathon, active Twitter coverage). Vendors would have preferred more engagement from attendees, but that’s generally the case with every event (virtual and physical).
  • Thyroid Ultrasound Cine Risk System: Stanford researchers developed a deep learning system that accurately classifies thyroid nodules using ultrasound cine images (versus static ultrasound images). The SIIM 2021 study (n = 192 nodules, 177 patients, 17,412 frames) showed that the cine ultrasound system classified benign and malignant nodules with a 0.858 AUC, which was higher but statistically similar to TI-RADS’ 0.798 AUC.
  • Disrupting NHS Diagnostics: A major NHS Group’s CEO publicly proposed expanding NHS imaging facilities into the country’s “high streets” (versus hospitals) and introducing new online self-service systems that place patients in control of their diagnostic exams (versus clinicians). Although much of this proposal hinges on future technology advancements (smaller scanners, AI efficiencies), the NHS is already starting to roll-out community diagnostic hubs, and more could be on the way.
  • A PI-RADS + PSAd Biopsy Strategy: A new Radiology Journal study found that a prostate biopsy decision strategy that forgoes biopsies among men with PI-RADS 3 lesions and PSA density below 0.1 ng/mL per milliliter is more effective than MRI-based risk models. The study (n = 385 men w/ mpMRI PI-RADS category ≥3) found that this PI-RADS + PSAd strategy would result in 63 fewer biopsies per 1000 men, without sacrificing detection of clinically significant prostate cancers.
  • Aetna Expands LDCT Coverage: Aetna officially adopted the USPSTF’s new lung cancer screening guidelines, lowering its starting age to 50 years (vs. 55yrs) and reducing its smoking history requirement to 20-pack years (vs. 30yrs). This is a major milestone for the new LDCT screening guideline, noting that Medicare and the other major private payers are currently evaluating their own lung cancer screening coverage.
  • Micro-X’s Mobile FDA: Micro-X’s Rover mobile X-ray system just gained full FDA approval, expanding upon the lightweight system’s FDA approval for military use. The new configuration uses a Varex-based digital Flat Panel Detector and imaging processing software system, while the military version uses Fujifilm technology.
  • Ultrasound Bone Age Scoring: Chinese researchers developed an ultrasound bone age scoring system (sum of radius, ulna, and femur ossification x 100) that rivals X-ray’s accuracy, without the radiation. The researchers performed the ultrasound bone age scoring method on 1,089 children (160 w/ suspected bone growth disturbances), diagnosing abnormal bone age with high sensitivity (93% boys, 100% girls) and specificity (98% boys & girls).
  • Cortical Thickness MRI for Dementia Progression: South Korean researchers found that MRI cortical thickness combined with clinical variables could help predict dementia progression among patients with Parkinson disease. The team developed a series of machine learning models, achieving the greatest accuracy with models using both cortical thickness and clinical variables (0.80–0.88 AUCs) versus models trained with only cortical thickness or clinical data (0.75–0.83 & 0.70–0.81 AUCs).
  • Zebra & Human Bytes’ Nordic Alliance: Zebra Medical Vision and Danish healthcare AI startup, Human Bytes, announced that Human Bytes will sell and market Zebra-Med’s public health AI solutions for osteoporosis and coronary artery disease in Europe’s Nordic countries. Human Bytes just launched in January, but it was founded by local medtech veterans and now has partnerships with Zebra-Med, Arterys, and Mirada.
  • Private AI: The team behind the PriMIA (Privacy-preserving Medical Image Analysis) deep learning system provided new details on its privacy safeguards and showed that it can accurately identify pneumonia in pediatric CXRs. PriMIA combines federated learning (shares algorithms across sites, not data) with additional privacy protection layers (anonymizes sources & patients, prevents reconstruction), which they believe could streamline AI development and help overcome legal/ethical barriers.
  • iThera Adds Ultrasound: iThera Medical’s second-generation MSOT Acuity Echo system, which combines Multispectral Optoacoustic Tomography (MSOT) with ultrasound tomography, just gained CE Mark certification. With the new system, clinicians can use tomographic ultrasound to identify relevant anatomical structures and use MSOT to examine tissue for functional and molecular biomarkers.

Oxford and Zebra’s VCF Detection Turnaround

Learn how the University of Oxford increased its vertebral compression fracture detection rate from 50% to 90% using Zebra Medical Vision’s bone health solution.

– Sponsored.

The Resource Wire

  • It’s clear that structured reporting is a must for CVIS platforms, but they aren’t all created equal. This Hitachi article reveals what physicians and sonographers view as the “non-negotiable” CVIS structured reporting features.
  • In this Novarad video, interventional oncologist Gary M. Onik, MD shares how Novarad’s AR surgical navigation system, OpenSight, helps his team accurately assess and treat tumors.
  • Did you know 80% to 90% of sonographers experience pain while performing scans at some stage in their career? Check out this Canon Medical Systems video detailing its latest innovations that improve sonographer comfort and help reduce risk of injury.
  • United Imaging’s “all-in” approach means that every system ships with its entire suite of features and capabilities (no options), giving its clients more clinical flexibility and cost predictability.
  • It says a lot when a solution works so well for a radiology department that they decide to perform a study to quantify its benefits. In this Imaging Wire Q&A, University Hospital of Zurich’s Thomas Frauenfelder discusses his experience and study on Riverain Technologies ClearRead CT.
  • This AI economics overview from Healthcare Administrative Partners details the various AI ROI scenarios and ways that AI can contribute to radiology practices until reimbursements become more of a reality.

You might also like

You might also like..

Select All

You're signed up!

It's great to have you as a reader. Check your inbox for a welcome email.

-- The Imaging Wire team

You're all set!