Uncategorized

CV19 Imaging Gets Real | Handheld micro-OCT | More AB-MRI Evidence

“The timing is inelegant.”

An unnamed lobbyist commenting on efforts to include anti-provider surprise medical billing legislation in the federal coronavirus package while providers are simultaneously fighting on the CV19 frontlines.


Imaging Wire Sponsors

  • Focused Ultrasound Foundation – Accelerating the development and adoption of focused ultrasound.
  • GE Healthcare – Providing point of care ultrasound systems, from pocket-sized to portable consoles, designed to support your clinical needs and grow along with your practice.
  • Healthcare Administrative Partners – Empowering radiology groups through expert revenue cycle management, clinical analytics, practice support, and specialized coding.
  • Nuance – AI and cloud-powered technology solutions to help radiologists stay focused, move quickly, and work smarter.
  • Qure.ai – Making healthcare more accessible by applying deep learning to radiology imaging .
  • Riverain Technologies – Offering artificial intelligence tools dedicated to the early, efficient detection of lung disease.

The Imaging Wire

CV19 Imaging Gets Real

The COVID-19 pandemic has become very real and so has CV19 imaging. News from the last few days revealed that we’re learning more about what’s working in CV19 radiology, showed how regulatory changes are making their way to imaging teams, and gave more evidence that healthcare’s CV19 focus is dragging down overall volumes. Cheers to the helpers.

  • CV19 CT Advice – A UPenn duo called upon radiologists to adapt to the COVID-19 pandemic, providing a collaboration-focused solution for those concerned about overcalling or undercalling potential CV19 patients. To avoid these extremes, they encouraged radiologists to build their familiarity of CV19 findings in CT scans and then work with the overall care teams to ensure accurate and balanced diagnosis.
  • NY Loosens Rules – The state of New York loosened up its healthcare provider laws to support the COVID-19 fight, including a number of changes that will impact radiology departments. The state: 1. Allowed licensed radiologic technologists from any state to practice even if they aren’t registered in NY; and 2. Gave physicians and a range of other healthcare providers immunity from civil liability for mistakes made during the COVID-19 response (except gross negligence).
  • CXR Inferior, But Usable – A new study in Radiology added further evidence that chest X-ray is inferior to RT-PCR tests and CT for COVID-19 diagnosis, but is still usable. In a study of 64 patients (58 w/ CV19), chest X-ray achieved 69% sensitivity (vs. RT-PCR’s 91%), and although it spotted similar abnormalities as CT, they were visible later after onset of symptoms with CXR (10-12 days vs. 6-11 days). That said, the researchers suggested that CXR can still play a role in initial CV19 screening, particularly as a way to reduce burden on CT units.
  • Mednax’s Warning – We’ve been hearing more and more about CV19’s negative impact on scan volumes (here’s one from last week). Mednax made CV19’s imaging impact even more real after revealing significant declines in its anesthesiology and radiology businesses due to delays in non-urgent medical procedures.
  • COVID-Net – Canadian AI startup DarwinAI and the University of Waterloo co-developed a chest X-ray-based open-source neural network for CV19 diagnosis (n = 5,941 chest X-rays, 2,839 patients) intended to help clinicians and scientists create CV19 diagnosis tools. The new COVID-Net CNN will later be coupled with an explainability tool that shows how their AI technology reaches its COVID-19 detection decisions.



Handheld micro-OCT

A Singapore-led team of scientists unveiled a new handheld micro-OCT device that could be used to allow clinicians without radiology or pathology expertise detect early signs of cancer.

  • The Tech – The device’s micro Optical Coherence Tomography (micro-OCT) tech emits near-infrared light that penetrates tissue and organs and measures the ‘echo’ from its light waves to create cross-section images as small as 1 to 2 micrometres. These images are processed by custom software, which transforms 2D cross-section images into a 3D picture and renders different parts in color, potentially showing the first signs of tumors in cells.
  • Applications – The researchers are initially targeting colon, stomach, and skin cancer diagnosis, which begin in the nuclei of epithelial cells, and are too small to image with conventional technology (1-2 micrometers vs. 0.5 millimeter). The micro-OCT device’s first trial revealed that it could detect abnormal colon polyps with 95% accuracy, equaling trained pathologists.
  • Users & Patients – The device was designed for medical professionals who do not specialize in imaging or pathology and for use at the bedside, while supporting patients who might not have timely access to advanced imaging equipment.
  • Next Steps – In addition to ongoing research, the new system is now being commercialized by a Chinese firm, Suzhou Sai Luo Er Medical Imaging Technology.



More AB-MRI Evidence

Abbreviated breast MRI (AB-MRI) is off to a good start in 2020. One month after an AB-MRI study garnered widespread headlines for its effectiveness among women with dense breasts, a new study came out highlighting its effectiveness with women with a personal history of breast cancer.

  • The Study – South Korean researchers retrospectively reviewed 1,880 screening AB-MRIs from 763 women with a history of breast cancer, along with the results of their tissue diagnosis or one-year follow-ups.
  • The Results – The 1st round AB-MRIs detected 15 of 21 cancer recurrences, achieving high sensitivity (100%), specificity (96%), and accuracy (96%), while the 2nd round of AB-MRI screenings caught six other cancers.
  • Conclusion – These 11-minute scans can improve cancer detection among women with breast cancer histories, given its high specificity, sensitivity, and PPV.

The Wire

  • Surprise Billing Surprise: A bipartisan group of federal lawmakers are reportedly working to include pro-insurer surprise medical billing legislation in the coronavirus package that would base benchmark rates on the average in-network prices within each geographical area. This could be seen as pretty harsh timing given the heroism of America’s doctors right now, but it’s also good timing for pushing through stalled legislation. As Winston Churchill (and Rahm Emanuel) once said, “never let a good crisis go to waste.”
  • True Theranostic: Australia’s ANSTO radioisotope researchers developed a new true theranostic agent, called Scandium-47, that could be used for targeted cancer therapy and diagnosis with SPECT imaging. Scandium’s potential as a true theranostic agent is unique, as theranostic approaches usually involve using two chemically similar radioisotopes for therapy and diagnosis.
  • Butterfly Disables: Butterfly Network disabled its Butterfly iQ ultrasound’s Auto Ejection Fraction Tool and the Auto Bladder Volume Tool because they were added to the device without FDA approval. Butterfly called the removals temporary and is working with the FDA to re-release the software tools, which are key parts of its AI-assisted value proposition.
  • Bad AI Science: A new paper in BMJ detailed how few medical imaging AI studies actually follow research best practice. Here’s their list of what AI research is often getting wrong: 1. Few studies are randomized (10 randomized vs. 81 non-randomized); 2. Few studies are prospective (9 of the 81 non-randomized trials); 3. Most studies have a high risk of bias (58 of the 81 non-randomized trials); 4. Data and code availability are often lacking (unavailable in 93%-95% of studies); and 5. Most human comparer groups are too small (median of 4 experts). Despite all of these issues, 61 of 81 studies still stated that their AI models performed comparable to (or better than) clinicians.
  • ACR’s RO Delay Request: The ACR called on CMS to delay the start of its new radiation oncology payment model until January 2021, arguing that the change wouldn’t allow radiologists to focus on treating CV19 patients. This isn’t the first time that the ACR tried to delay the new RO payment model (this is just a more noteworthy reason), as it’s concerned that the model’s prospective payments structure might have a negative financial impact.
  • EWI US for Arrhythmia: Columbia University researchers found that electromechanical wave imaging (EWI) ultrasound can identify the location of arrhythmias more effectively than standard 12-lead electrocardiograms (ECG). In a double-blinded study, six electrophysiologists located arrhythmias in 55 patients with cardiovascular disease with 96% accuracy using EWI versus just 71% using ECG.
  • Telerad AI Growth: A new post from Signify Research outlined the drivers of teleradiology’s expected growth (e.g. radiologist shortages, growing advanced imaging volumes, new regulations) and suggested that AI will play an increasing role in teleradiology given AI’s contribution to speed, accuracy, and workflow prioritization. Signify doesn’t expect AI to have a major impact on teleradiology within the next five years due to a variety of factors (e.g. slow regulatory process, shortage of studies and annotated datasets, slow IT integrations), but it does expect telerad AI growth in the long run.
  • Butterfly Migrates to Canada: Health Canada approved Butterfly Network’s Butterfly iQ ultrasound system, making it available to Canadian healthcare practitioners and health systems for the first time. The launch announcement overwhelmingly focused on Butterfly iQ’s ability to help monitor and triage CV19 patients, which is certainly relevant given the growing pandemic, but represents a change from Butterfly’s traditional whole body messaging.
  • Ethical AI Framework: A Stanford team unveiled a new framework for sharing clinical data for use in AI development. The framework suggests that: 1. Nobody truly owns clinical data (not patients, nor institutions), so patients might not have to consent to data sharing as long as their privacy is protected; 2. Clinical data should be made available to researchers / developers after it’s aggregated / anonymized; 3. Everyone who accesses this data should be held to high ethical standards (protecting privacy, not selling clinical data); 4. Providers could share clinical data with industry partners who financially support their research, if the support is for research and not just for the data.
  • McKesson’s GBCA Case Dismissed: McKesson was able to get most of a Gadolinium Deposition Disease lawsuit dismissed. A federal court found that because McKesson was a distributor for its Omniscan MRI agent (not the manufacturer) it isn’t liable for potential product risks or responsible for labeling to communicate the risks.
  • ML for MD: Spanish researchers developed a machine learning tool that helps doctors diagnose muscular dystrophy by analyzing patterns of muscle fatty replacement in muscle MRI scans. The team used 976 pelvic and lower limb T1 muscle MRIs from 10 different MDs to create 2,000 different models that quantify fatty replacement. The best of the models achieved 95.7% accuracy, 92.1% sensitivity, and 99.4% specificity against a 20-MRI test set (with higher accuracy than four experts).
  • EchoNous KOSMOS FDA Approved: Just days after landing CE Mark Approval, EchoNous announced the FDA approval of its KOSMOS Platform, which includes an 8 oz handheld ultrasound system and its new AI-enabling KOSMOS Bridge tablet. The KOSMOS AI platform can currently calculate systolic heart function measurement (Left Ventricular Ejection Fraction), while EchoNous plans to launch a trio of algorithms focused on improving users’ learning / confidence curves and labeling heart anatomy.

The Resource Wire

  • In this GE Healthcare video, ultrasound users and educators discuss how the Vscan Extend handheld ultrasound combines portability and intuitive design so you can use it in the moment to potentially change patient outcomes.
  • This Nature Research study found that Qure.ai’s qXR tool can identify TB-associated abnormalities in chest radiographs with over 95% accuracy, potentially eliminating 2 out of 3 Xpert MTB/RIF tests.
  • Healthcare Administrative Partners detailed how relaxed telemedicine rules during the COVID-19 Public Health Emergency could create an opportunity for radiologists to use telemedicine in their practice in this blog post.
  • Riverain Technologies is dedicated to providing enterprise software tools to aid clinicians in the efficient, effective, early detection of lung disease. Learn more.
  • Nuance is providing its healthcare customers with free add-on licenses to help them scale up during the COVID-19 emergency, including 90-day add-ons to its PowerShare and PowerScribe solutions.
  • This article details ongoing efforts to close the reimbursement gap for high intensity focused ultrasound prostate cancer treatment.

You're signed up!

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

-- The Imaging Wire team