#367 – The Wire

  • HiP-CT Revolution: UCL scientists unveiled the first hierarchical phase-contrast tomography (HiP-CT) images of intact human organs, showing that HiP-CT scans can produce a diverse range of clinical views (whole-organ, functional, blood vessels, cellular). HiP-CT is getting headlines for its impressive technology (its 100b-times brighter than current X-rays), but the UCL team seems more excited by HiP-CT’s ability to zoom-in on the smallest organ features, suggesting that it might “revolutionize” diagnostic processes and even our understanding of human anatomy.
  • PowerShare Image Aware: Nuance and Collective Medical joined forces to launch PowerShare Image Aware, a new solution that provides ED physicians with their patients’ prior radiology studies in real time and within existing clinical workflows. Nuance PowerShare Image Aware also shows if images and associated metadata are available in PowerShare for viewing, while allowing radiologists to integrate a patient’s prior imaging into radiology reporting tools such as Nuance PowerScribe.
  • COVID Burnout: A survey of 3,176 Sutter Health clinicians found that 29% experienced burnout during the COVID-19 pandemic (as of mid-2020), revealing higher likelihoods of burnout for women (odds ratio = 2.19), clinicians with childcare/caregiving responsibilities (OR = 2.19), and clinicians in emergency medicine and radiology settings (ORs = 1.58 & 1.87). Surprisingly, the report called radiology’s high burnout rates “surprising” and theorized that it might be because the pandemic reduced their “work or income,” showing that many healthcare experts aren’t aware of radiology’s long history of burnout (that’s definitely not due to reduced work).
  • Hyperfine’s Gates Grant: Hyperfine received a $3.3m grant from the Bill & Melinda Gates Foundation (following a previous $1.6m grant) that will fund the deployment of 25 Hyperfine Swoop portable MRIs in developing countries. The Swoop systems will be used to study the neurological effects of malnourishment on young children, while also assessing the Swoop’s feasibility for pediatric neuroimaging in low-resource settings. 
  • Self-Aware AI: Google researchers developed a dermatology classifier AI tool that “knows what it doesn’t know” when it’s used to diagnose a condition that wasn’t included in its training data. This paper is pretty complex if you don’t have an AI background, and it isn’t about our type of medical imaging, but it’s worth a look for readers trying to understand how to handle outlier conditions.
  • A Case for Community Multispecialists: A new AJR study provided more evidence in favor of staffing multispecialty radiologists in acute community settings. The researchers analyzed 5.9m acute exams from community hospitals that featured preliminary interpretations from specialist telerads and final interpretations from on-site radiologists. Interpretation discrepancy rates didn’t increase when the telerads read exams from outside of their subspeciality, but did increase when telerads interpreted “advanced” exams from within their subspeciality (0.26% vs. 0.18% major discrepancy rates; 0.34% vs. 0.29% minor discrepancy rates).
  • PaxeraHealth’s DIY AI: PaxeraHealth introduced its forthcoming “Ark” zero-coding AI authoring platform, which healthcare systems and academic centers could use to “DIY their own AI algorithms.” Providers will be able to utilize collaborative learning to train AI models leveraging data from other Ark members, expediting imaging AI production.
  • deGrom’s MRI Injury: Ace New York Mets pitcher, Jacob deGrom, blames the elbow injury that ended his 2021 season on a long and awkwardly positioned MRI exam. The exam required deGrom to lie on his stomach with his elbow raised above him for nearly an hour, and revealed only a mild forearm strain. However, deGrom’s discomfort spread to his elbow during the next few days, and he wasn’t cleared to play again until the season was nearly over. Given deGrom’s $33.5m annual salary, this might be the most expensive MRI exam ever.
  • ED Errors and Adverse Outcomes: When emergency imaging diagnostic errors cause patients to return to the ED for more imaging, it often results in adverse medical outcomes. That’s from a South Korea-based study that analyzed 1,054 patients who received emergency MRIs and CTs and returned to the ED for additional imaging within 7 days (231 due to diagnostic errors), finding that patients who had diagnostic errors experienced far higher rates of adverse outcomes within the next 30 days (33.3% vs. 14.8%).
  • Covera & HealthHelp: Covera Health revealed plans to integrate its national Centers of Excellence program (CoE) into HealthHelp’s Diagnostic Imaging program. This seems like a logical combination, as providers use HealthHelp’s Diagnostic Imaging program to ensure appropriate imaging referrals and providers/plans use Covera Health’s CoE program to ensure that their patients are referred to quality imaging practices. It also significantly expands Covera’s potential CoE user base.
  • Brain CT DL, With and Without Contrast: Japanese researchers developed a deep learning model that used a combination of contrast-enhanced and non-contrast CT head images (CE+NECT) to improve brain metastases detection. The team developed the CECT+NECT model using scans from 116 patients with brain metastases (428 metastases for training/validation, 231 for testing), which outperformed a contrast-enhanced CT model for sensitivity (88.7% vs. 87.6%), PPV (44% vs. 37.2%), and false positives per patient (9.9 vs. 13.6). 
  • CyncHealth Selects PowerShare: CyncHealth selected Nuance PowerShare as the Iowa and Nebraska health information exchange’s patient imaging sharing platform, including Nuance’s new PowerShare Image Aware solution that provides ED physicians with patients’ prior imaging.

You're signed up!

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

-- The Imaging Wire team