#371 – The Wire

  • An LDCT Coverage Milestone: After a lengthy review, CMS found “sufficient” evidence for expanding Medicare’s lung cancer screening coverage to younger people with shorter smoking histories (50-77 vs. 55-77 age range; 20 vs. 30 smoking pack-years), aligning with the USPSTF’s early 2021 recommendations. This is a key step towards expanding public coverage for LDCT screening, which could become official after an upcoming comment period. It could also lead to similar coverage expansions by private plans. 
  • GE’s PCCT Study: GE Healthcare launched a research collaboration with Karolinska Institutet and MedTechLabs to clinically evaluate and optimize GE’s forthcoming Photon-Counting CT systems. The pilot study will focus on the performance of GE’s pure silicon CT detector technology, which GE acquired from Prismatic Sensors last year, and reportedly has both clinical and manufacturing advantages. We don’t usually cover studies before they happen, but this one is notable given Photon-Counting CT technology’s clinical potential.
  • More CAC AI Evidence: A new study out of China detailed an AI CAC scoring model (AI-CACS) that was able to quickly and accurately produce CAC scores using non-gated chest CTs, continuing AI-based CAC scoring’s recent momentum. The researchers used the AI-CACS software to analyze non-gated / non-contrast chest CTs from 901 patients (who also underwent gated cardiac CTs), producing CAC scores that were very similar to manually-generated CAC scores from the gated cardiac CT exams.
  • Sirona & RevealDx: Imaging informatics startup Sirona Medical and AI startup RevealDx announced plans to integrate RevealDx’s lung nodule decision support software into Sirona’s unified Radiology Operating System (RadOS). The alliance brings the first AI partner to Sirona’s forthcoming RadOS platform while expanding RevealDx’s future platform presence beyond its alliance with Volpara.
  • Philips Ultrasounds Quantify Liver Fat: Philips announced the FDA 510(k) clearance of its Liver Fat Quantification solution, making it available with its EPIQ Elite and Affiniti ultrasound systems. The new Liver Fat Quantification solution improves clinicians’ ability to catch early-stage fatty liver disease, assess severity, and track longitudinal changes, all of which were difficult with traditional grayscale imaging. 
  • Fast Quantitative Low-field MRI: A new study out of Switzerland highlighted how low-field MRI could benefit from multiparametric technology, potentially improving both scan speed and clinical value. The researchers developed a fast multiparametric low-field MRI (0.1T) that imaged a human hand and wrist in just 8.5 minutes, suggesting that rapid quantification could expand point-of-care MRI to new metrics and contrasts.
  • Owkin’s Unicorn Status: French startup Owkin joined the exclusive group of imaging AI “unicorns” (>$1b valuation), after landing an $180m investment from pharma giant Sanofi. Owkin has done some interesting clinical imaging AI research. However, Owkin is different from the AI firms we usually cover, as it uses federated learning to analyze a range of medical data (not just imaging), and its business model largely focuses on supporting drug discovery and life science research (explaining Sanofi’s interest).
  • Orthopedic AI Evidence: A new Skeletal Radiology study assessed ImageBiopsy Lab’s knee alignment DL model, finding that it was able to accurately perform assessments and could help automate this time-consuming process. The researchers used the commercially available DL model to analyze 295 long-leg radiographs (LLRs), matching manual annotations with 89.2% accuracy and achieving high reproducibility, while performing each assessment an average of 130 seconds faster than clinicians.
  • Advantis’ Brain MRI FDA: Advantis Medical Imaging announced the FDA clearance of its Brainance MD brain MRI analysis platform. The browser-based solution supports the processing, visualization, and analysis of a range of brain MR imaging techniques (DTI, DSC perfusion, fMRI), reportedly reducing manual work and supporting interpretation.
  • AI Efficiency: A prospective study out of Denmark showed that integrating AI into PACS reading workflows doesn’t necessarily extend radiologist interpretation times. In the study, a radiology resident and a thoracic radiologist interpreted non-contrast low-dose chest CTs with and without Siemens Healthineers’ AI Chest Companion (n = 20 with & 25 without). The AI tool had a statistically insignificant effect on average reading times (Radres = 437s w/ AI vs. 370s without; Rad = 380s vs. 366s), while reader perceptions were very different (Radres felt AI added time to 50% of cases; Rad = just 5%).
  • PillCam FDA: Medtronic’s unique PillCam gained FDA 510(k) clearance for at-home endoscopy procedures. The now-approved PillCam SB3 @HOME kits are delivered directly to patients, allowing gastroenterologists to remotely examine patients’ small bowels. Although not technically our kind of imaging, the momentum we’re seeing in the capsule endoscopy arena is notable, with other options coming from Check-Cap (low-dose X-ray) and InsideOut (videos).

You're signed up!

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

-- The Imaging Wire team