“A little knowledge, so the saying goes, can be a dangerous thing.”
Signify Research’s Alan Stoddart on the potential pitfalls of ultrasound simplification and home adoption.
Know anyone who made outstanding contributions to radiology practice and outcomes this year? Nominate them for an Imaging Wire Award.
Imaging Wire Sponsors
- Bayer Radiology – Providing a portfolio of radiology products, solutions, and services that enable radiologists to get the clear answers they need.
- GE Healthcare – Enabling clinicians to make faster, more informed decisions through intelligent devices, data analytics, applications and services.
- Healthcare Administrative Partners – Empowering radiology groups through expert revenue cycle management, clinical analytics, practice support, and specialized coding.
- Hitachi Healthcare Americas – Delivering best in class medical imaging technologies and value-based reporting.
- Nuance – AI and cloud-powered technology solutions to help radiologists stay focused, move quickly, and work smarter.
- Riverain Technologies – Offering artificial intelligence tools dedicated to the early, efficient detection of lung disease.
- Siemens Healthineers – Shaping the digital transformation of imaging to improve patient care.
The Imaging Wire
AI Faceoff
Most folks in imaging AI have been trying their best to avoid the controversial /clichéd “AI vs. clinician” angle, but a new JAMA study brought it back into the headlines. More importantly, it provided new evidence that AI can enhance radiology workflows:
- The Study – The IBM and USC research team developed an AI algorithm (trained on 342k frontal CXRs) to analyze 72 different frontal CXR findings. They then compared the model’s performance analyzing 1,998 CXRs against reads from five 3rd year radiology residents.
- The Results – The algorithm and residents had comparable sensitivity (0.716 vs. 0.720 AUCs), while the model achieved higher specificity (0.980 vs. 0.973) and positive predictive values (0.730 vs. 0.682).
- The Significance – These results suggest that a well-trained AI algorithm could effectively perform preliminary interpretations, helping improve efficiency, accuracy, and cost of care. It could also help improve access in regions that don’t have enough radiologists.
- What about the “AI vs. clinician” Angle? – Many imaging AI studies compare model performance to either blinded readers or previous interpretations made by human rads, so this study isn’t all that unique in that regard. That said, any algorithm that would handle most initial interpretations and has a goal to “reduce costs” is also positioned to take read volumes away from human radiologists.
The Wire
- COVID in the CT: An Italian team inspected the internal gantry components of their CT machine for signs of COVID-19 after scanning 180 COVID-positive patients in March, finding SARS-CoV-2 RNA in the CT’s inward airflow filter but no other components (e.g. case, motor, tube, outflow tube, etc.). This could actually be good news, suggesting that the inward airflow filter “could be a barrier to SARS-CoV-2 dissemination.”
- Weill Cornell’s Radres Ultrasounds: As part of its expanded ultrasound education program, Weill Cornell distributed a Philips Lumify portable ultrasound to each of its radiology residents. Although still unique enough to get the attention of many folks on Radiology Twitter, med school handheld ultrasound handouts like this seem to be increasingly common. Everyone from UC Irvine Medical School’s incoming class got their own Butterfly iQ in August 2019 and Indiana University’s anesthesia program gave its residents Butterfly iQs in January.
- COVID Severity Predictor: CT severity score combined with clinical data can be used to accurately predict whether a newly-admitted patient’s COVID-19 symptoms will become serious. That’s from a new study out of China (n = 247 w/ moderate symptoms) that found patients with high CT severity scores (1.32 odds ratio), crazy-paving CT signs (3.09 OR), hypertension (4.93 OR), higher NLR ratios (2.13 OR), and advanced age (1.09 OR) have the greatest risk of developing severe symptoms. Using that data, the researchers created a risk model using CT score, age, and NLD to predict severity risk with a 0.867 to 0.898 AUC.
- Signify on Home Ultrasound: A new post from Signify Research forecast major handheld ultrasound growth (Signify: ~$225m revenue in 2020 to ~$460m in 2024), suggesting that wider hospital and practice adoption will drive most of this growth, but home ultrasound adoption could contribute. As the COVID-19 pandemic normalizes telemedicine, we’re starting to see handheld ultrasound players target the at-home market (e.g. Butterfly & PulseNmore) with tools to support these untrained users (e.g. teleguidance, automation tools, how-to videos), and with costs that are becoming more justifiable for payors or even some personal users (Signify: ~$6k avg. 2017 to ~$3.9k avg 2024).
- Prostate mpMRI Benefits: New research out of Australia found that the country’s move to fund multiparametric MRI prostate (mpMRIp) screening in 2018 reduced Australia’s prostate biopsies by 354.7 per month and cut annual biopsy-related costs by $9.4m (USD).
- Cigna’s Site of Care Limits: After a brief COVID-related delay, Cigna officially launched its High-tech Radiology site of care policy that limits the circumstances that in-hospital high-tech imaging exams (MRI/MRA/CT/CTA) are covered. Cigna outlined a number of patient situations where onsite high-tech imaging is “medically necessary” (e.g. patients <10yrs, obstetrical observations, perinatology services, transplantation services, patients w/ contrast allergies, preoperative imaging, and when there offsite imaging isn’t an option). Patient situations deemed “not medically necessary” will only be covered at standalone imaging centers.
- Imbio & Genentech: Imbio announced a partnership with biotech heavyweight Genentech to develop quantitative imaging diagnostics for lung diseases. The alliance will combine Imbio’s imaging AI expertise (technical, regulatory, and commercial) with Genentech’s pulmonary disease expertise to develop AI that could enhance drug development and deliver more personalized care.
- fMRI Levels Off: Emory University researchers found that fMRI use among Medicare patients grew significantly between 2007 and 2014 (17.7 per 1m enrollees to 32.8 per 1m) after the introduction of an fMRI CPT code, but utilization has been flat since then. The researchers suggested that the drop-off in fMRI growth might be due to reduced reimbursements (from $618 to $455), the increased adoption of magnetoencephalography, and fMRI’s many requirements (tech, time, and expertise).
- Private Prenatals: Expectant parents in Canada are increasingly using (and paying) private ultrasound clinics for their prenatal exams during the COVID-19 pandemic in response to a public hospital policy that keeps dads and other family members out of the exam room.
- Shimadzu’s RADspeed Pro V: Shimadzu Medical Systems USA unveiled its new RADspeed Pro V general radiographic system. The new model adopts most core features found in Shimadzu’s other RADspeed Pro systems, while adding support for three panel sizes (14×17, 17×17 and 10×12 inches), an improved detector (more durable, 16hr battery life), and support for heavier patients (880lbs. max load capacity).
- CEM Targeted Breast Ultrasound: A new AJR study detailed the value of using targeted breast ultrasound for characterizing and potentially guiding biopsies on lesions initially identified with contrast-enhanced mammography (CEM). The researchers reviewed 1,000 consecutive CEM examinations with same-day targeted breast US exams, finding that 31% of the lesions detected with CEM (n = 153 lesions) correlated with the targeted ultrasound results. These CEM/US correlated lesions were far more likely to be malignant that those only identified with CEM (26% vs 10%), while ultrasound-guided biopsies after the CEM-targeted US exams had a PPV of 32%. The researchers suggest that CEM/US could reduce follow up MRI exams and MRI-guided biopsies.
- Nanox Patent Spotlight: A deep dive from patent and IP specialists, Rutman IP, poked more holes in Nanox’s “breakthrough” technology claims, suggesting that Nanox’s X-ray source might not be superior or more economical than current options and it’s patent coverage “does not appear to be strong.” Among Rutman IP’s numerous criticisms, the report suggests that: 1) Nanox’ tube cost advantage comparisons are “apples-to-oranges;” 2) There’s no patent-based evidence that the Nanox tubes perform better; 3) Nanox’s lack of lifetime and stability data suggest that the technology might still be at the “conceptual stage.”
- Six Security Steps: A new report in Academic Radiology detailed the steps hospitals should take to protect their imaging infrastructure from cyberattacks. Here’s what they suggest hospitals do: 1) Deploy virus scanners on all possible/permitted imaging systems; 2) Systematically manage software updates; 3) Clarify security update process/schedule and implementation plan with device vendors; 4) Create a long-term access rights strategy with PACS vendors; 5) Have a clear security policy for all devices and software; 6) Use encryption and authentication wherever possible.
- Philips & LeQuest Simulate Training: Philips announced an ultrasound training partnership with LeQuest, allowing Philips Affiniti users to use LeQuest’s interactive training platform to practice with a range of simulation scenarios (at any time, without requiring device access). LeQuest will provide training modules for all Affiniti models and each clinical segment (general imaging, obstetrics/gynecology, and cardiology), with training modules initially launching in Latin America and parts of Europe before expanding to other regions.
The Resource Wire
– This is sponsored content.
- In this Nuance video, Penn Medicine professor, Warren B. Gefter, shared how PowerScribe One leverages AI, structured data, and automation to drive improved patient care.
- This Bayer Radiology case study details how its Certegra P3T Software automates contrast enhanced abdominal CT injection protocols based on patient characteristics and contrast concentration.
- Did you know that one in three Americans is obese and obesity is even more prevalent in rural communities? This Hitachi blog shows how its wide aperture CT and MRI systems are the best fit for rural hospitals, helping them care for patients of all sizes and get more ROI from their imaging systems.
- They say that in times of crisis, you get to know who your real friends and partners are. This Q&A session details how Healthcare Administrative Partners stepped up to guide their client Triad Radiology Associates through the challenges presented by the COVID-19 pandemic.
- This GE Healthcare article details how Big Sur’s Mee Memorial hospital leveraged AI-powered POCUS and X-ray technologies as wildfires and the COVID-19 pandemic brought in a new wave of patients.
- Learn how Riverain’s ClearRead CT Vessel Suppress provides a powerful and intuitive view for clinicians through the suppression of vascular structures.
- Check out how Siemens Healthineers’ Tin Filtration is like selecting better quality sunglasses.
