fbpx

Wellness Ultrasound | Automatic Labels | Orthopedic Reasons

“Burns can be expected in MR.”

One radiographer’s reason for not reporting burn-related MRI incidents.


Imaging Wire Sponsors

Arterys | Bayer Radiology | Canon Medical Systems
Fujifilm Healthcare Americas | GE Healthcare |
Healthcare Administrative Partners | Novarad | Nuance
Riverain Technologies | Siemens Healthineers
United Imaging | Zebra Medical Vision



The Imaging Wire


Adding Ultrasound

A new Journal of Ultrasound in Medicine study found that ultrasound exams might be a valuable addition to Medicare Wellness visits, and could significantly increase detection and diagnosis.

  • The Background – Medicare Wellness visits are supposed to supplement annual physicals, so they generally only include a limited exam (blood pressure, weight, etc.) and aren’t considered useful by many patients and physicians.
  • The Study – Researchers from Minneapolis’ Abbott Northwestern Hospital performed ultrasound exams during 108 patients’ Medicare Wellness visits (65-85yrs), and measured abnormality detections, diagnoses, and follow-ups.
  • Detection & Diagnosis – The added ultrasound scans detected abnormalities in 94% of the patients (283 total abnormalities) and led to 172 new diagnoses, including previously undiagnosed chronic conditions.
  • Positive Impact – Adding ultrasound positively benefited 63.9% of the patients (only 1.8% were negative), while creating follow-up costs for 24% of the patients (15% <$50). Follow-up costs were lowest when PCPs were also ultrasound experts.
  • The Takeaway – We’d need a lot more evidence and POCUS training to justify adding ultrasound exams to the millions of Medicare Wellness visits performed in the US each year, but these results are notable and add to the growing field of evidence supporting primary care ultrasound.



ClearRead CT’s Impact at Einstein Medical

This Riverain Technologies case study details how Einstein Medical Center adopted ClearRead CT enterprise-wide (all 13 CT scanners) and how the solution allowed Einstein radiologists to identify small nodules faster and more reliably.

– Sponsored.


Novarad Simplifies Ditching the Disk

Do your patients text more than they use CDs? Find out how Novarad’s CryptoChart simplifies image access, combining secure QR codes and text and email communications to help providers and patients ditch the disk.

– Sponsored.



The Wire

  • Automated Labels: A King’s College London team developed a deep learning system that accurately labeled 121k brain MRI exams in under 30 minutes (versus “years”), representing a key step towards overcoming imaging AI’s labeling constraints. The researchers developed the system with 5k reference-standard labeled images (3k labeled as normal / abnormal, 2k defining abnormalities), which accurately classified the 121k test images as normal / abnormal (AUC-ROC = 0.991), and accurately defined the abnormalities (AUC-ROC > 0.95).
  • Volpara Adds Invitae: Volpara Health announced a partnership with Invitae that will make Invitae’s genetic testing services available in the US through the Volpara Breast Health Platform. The Invitae alliance is Volpara’s latest of many moves to expand its breast health platform and/or channel over the last few years, including a number of major acquisitions (MRS Systems, CRA Health) and partnerships (DetectED-X, Fujifilm, GE, Ambry, Screenpoint, MeVis, Myriad Genetics).
  • Do Orthopedic Surgeons Read Your Reports? Radiologists might be concerned that orthopedic surgeons never consult their reports, but a German study (n = 81 OSs, 27 departments) found that just 20% “never” read radiograph reports, only 4% “never” read CT reports, and 0% “never” read MRI reports (they routinely read = 43%, 67%, and 86%). When orthopedic surgeons skip radiology reports it’s because they weren’t available soon enough (24%), the surgeons didn’t have enough time (19%), or the reports had too much text (17%).
  • AI Certificates: It appears that RSNA and the University of Illinois were listening to the recent calls for more AI education. RSNA’s new Imaging AI Certificate Program will help radiologists (including less tech savvy rads) understand how to apply AI to their practices. Similarly, the U of Illinois’ new AI in Medicine Certificate program is intended to help clinicians without coding skills understand how AI works and how it can be applied.
  • The Annual Screening Advantage: A major breast cancer screening study out of Canada found that women with dense breasts have far fewer interval cancers if they undergo annual mammography screening. The study analyzed 148,575 women with dense breasts between 2008 to 2010 (77k w/ biannual screenings, 70k annual), finding that women who underwent biannual screenings were 63% more likely to develop cancer between screenings than those who attended annual exams (1.45/1k vs. 0.89/1k incidental cancer rates).
  • Infervision’s Big Series D: Chinese imaging AI company Infervision reportedly wrapped up a $139m Series D2 round (total now ~$213m) that it will use to fund R&D and expand its team, channel, and international presence. Infervision’s massive funding numbers might come as a surprise to readers focused on N. America and Europe, although Infervision has traditionally been among the AI funding leaders and Chinese AI firms have consistently landed large funding rounds over the last two years.
  • SPECT AI Effective: A new JACC study detailed a SPECT MPI-based deep learning model that was able to detect obstructive coronary artery disease (CAD) significantly more accurately than standard methods (in less than 12 seconds each). Against internal and external test sets (n = 3,578 & 555 patients), the model detected obstructive CAD more accurately (0.83 & 0.80 AUCs) than stress TPD tests (0.78 & 0.73) and reader diagnosis (0.71 & 0.65).
  • Voice Diagnosis: Voice symptom detection startup, Sonde Health, announced an alliance with Qualcomm that would make its vocal biomarker technology available with Snapdragon-based mobile and IOT products. Sonde Health definitely doesn’t utilize imaging, but its technology is intended to detect a range of imaging related respiratory issues (e.g. COPD, pneumonia, COVID), and it’s another example of how consumer mobile devices might serve as a future diagnostic starting point.
  • Generalists Match Specialists for Appendicitis: A new University of Wisconsin study revealed that general radiologists from community hospitals can diagnose appendicitis as accurately as fellowship-trained abdominal radiologists from academic hospitals, even when interpreting MRI exams. The study had three generalists and three specialists diagnose 198 patients (64 w/ appendicitis), revealing statistically similar average sensitivity and specificity for MRI exams (93.8% & 88.8% vs. 96.9% & 89.6%) and CT exams (96.9% & 91.8% vs. 98.4% & 93.3%). The study suggests that MRI is an effective alternative to CT in community-based generalist radiology practices.
  • MONAI Labeling Assistant: King’s College and NVIDIA’s MONAI group released MONAI Label v0.1, an open-source medical image labeling and learning tool that speeds up the creation of annotated datasets and helps AI developers continuously adapt their AI models. MONAI Label continuously learns from user interactions and training data, using that feedback to power “an AI-assisted annotation experience.”
  • Underreported MRI Incidents: A new European Radiology study found that just 38% of MRI incidents across a large Swedish region were officially reported between 2014 and 2019, including some catastrophic incidents. The study of 13 MRI units also found that the level of staff knowledge and number of MR physicists per scanner had a direct correlation with the number of annual incidents per scanner.
  • The Google Data Engine: Google Cloud unveiled its Healthcare Data Engine, which supports interoperability across healthcare data sources (medical records, claims, trials, research), gives clinicians a longitudinal view of patient records, and provides a cloud-based environment for analytics and AI.
  • CAD-RADS Advantage: A new Radiology Journal study found that CAD-RADS predicts major adverse cardiovascular events (MACE) among ED patients with acute chest pain more accurately than coronary artery calcium scores (CAC). Analysis of 1,492 patients (103 who experienced MACE), revealed that patients with the highest CAD-RADS categories were far more likely to experience MACE during a 31.5-month median period than patients with the highest CAC scores (8.5 vs. 4.4 hazard ratios).

Siemens’ Dual Source Cardiac CT

Explore untapped potential in dual source CT for cardiac imaging. This new Siemens Healthineers whitepaper showcases the clinical benefits of fast native temporal resolution.

– Sponsored.



The Resource Wire

  • See how GE Healthcare and SOPHiA GENETICS are pairing AI and genomics to help clinicians fight cancer in this new GE Healthcare article.
  • In this quick video, Sharp Memorial Hospital interventional radiologist, Jim Lyon, MD, describes the image quality and dose advantages of Canon Medical Systems’ Alphenix Sky+ system.
  • This Journal of Neuroimaging study found that 4D Flow MRI effectively supports cerebrovascular vasoreactivity analysis for extracranial‐to‐intracranial bypass planning and post-surgery evaluation, revealing a new use case for Arterys’ Cardio AI software.
  • 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.
  • Why United Imaging’s MI (uMI)? Every United Imaging molecular imaging system features its “uEXPLORER Inside” technology platform, which is designed for total-body scanning, is scalable for clinical systems, and excellent in an MR environment – you’ll see a big difference and your patients can benefit from their focus on coverage, clarity, and sensitivity.
  • Check out this Imaging Wire Q&A, where Bayer Radiology’s Dennis Durmis and MITA’s Peter Weems discuss the medical device service debate and how ongoing legislation and regulation efforts could impact patients, clinicians, and OEMs.