“It’s equally important to figure out why an ‘AI model’ can beat a human. Not just to help improve AI but rather to also improve the ‘human’.”
A Twitter comment from radiologist Erwin Carpio, MD.
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.
- 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.
The Imaging Wire
CXR Brixia Score
There’s been plenty of coverage of chest X-ray’s inferiority to CT for COVID-19 imaging, but CXR’s advantages (availability, cost, ICU applications, day-to-day monitoring) still helped drive the modality’s relatively wide adoption for CV19 management. With that in mind, a pair of Italian radiologists developed a chest X-ray scoring system (named the Brixia score) to help develop best practice for quantifying and monitoring COVID-19 progression with CXR.
- The Scoring Process – The Brixia scoring process involves dividing the lungs into six zones (two upper, two mid, two lower) on a frontal CXR and then scoring each zone on a 0 to 3 scale (18 = most severe). Zones with no lung abnormalities are scored zero, zones with interstitial infiltrates are scored one, zones with interstitial and alveolar infiltrates and interstitial predominance are scored two, and zones with interstitial and alveolar infiltrates and alveolar predominance receive a score of three.
- The Study – A sample of 100 reports from hospitalized CV19 patients using the 0-3 scoring method were reviewed by a thoracic radiologist. The hospital’s first readers and thoracic radiologist second reader achieved similar scoring ranges (0-15 vs. 0-16) and median scores (6.5 vs. 7), while both readers scored patients who died higher than patients who were discharged (p ≤ 0.002).
A team of Australia and UK-based scientists developed a new terahertz radiation-based imaging technology that could improve cancer detection, while reducing patients’ exposure to harmful radiation. Here’s how:
- The Tech – The devices’ owe much of their “potential breakthrough” status to their use of terahertz radiation (does not damage living tissue) and construction from nanowires (extremely thin). These technologies could “help create a new safe imaging technology with much higher resolution than current ultrasound devices.”
- Terahertz Advancement – Terahertz technology is already in use in airport security systems as well as industrial and clinical applications. However, most current terahertz systems can only detect radiation intensity, while this new technology can detect other properties, including polarization.
- Clinical Application – The terahertz imaging devices could be used to detect small tumors during check-ups and potentially spot suspicious tumors based on their specific terahertz ‘fingerprint.’
- X-Ray Growth: A new Informa Tech post forecast that the COVID-19 emergency will drive a 13% to 35% increase in digital radiography system unit sales in 2020 (originally forecasted +9% before CV19). As expected, this growth will benefit diagnostic X-rays (especially mobile), while interventional X-ray systems and C-arms will see big declines this year.
- ALZ Risk AI: A Boston University-based team developed a deep learning model that uses brain MRI scans along with other inputs (age, gender, mini-mental state exam scores) to estimate Alzheimer’s disease risk. The team trained the model on a dataset from 417 subjects (either w/ Alzheimer’s or cognitively normal) and validated it on three independent cohorts (n = 382, n = 102, n = 582), achieving AUCs between 0.876 and 0.99 (0.95 avg.) and surpassing a team of eleven neurologists (0.920).
- CV19’s $202B Cost: A new AHA report estimates that U.S. health systems and hospitals will absorb $202.6 billion in losses between March and June due to the costs associated with treating CV19 patients (-$36.6b), the effects of cancelled services ($161.4b), and protecting and supporting frontline staff ($4.6b).
- UHC Prioritizes CTA & FFRct: UnitedHealthcare joined the growing list of major insurers prioritizing coronary CTA and FFRct over standard cardiac stress testing (standard = nuclear stress, stress echocardiogram, PET Stress) among patients with low and intermediate risk of coronary artery disease.
- TeraRecon’s New CV19 Tools: TeraRecon rolled out a pair of new lung density analysis solutions intended to help assess and manage CV19-positive patients. The new LDA-II for Intuition allows existing TeraRecon Intuition visualization software users to add custom lung workflow. TeraRecon’s new Emergency Lung AI Suite is a cloud-based platform of lung density analysis tools intended for quick-deployment during urgent situations.
- PA & NC’s Liability Shields: Pennsylvania and North Carolina became the latest states to protect healthcare workers (including radiologists) from liability while treating patients during the COVID-19 emergency. The new laws come a month after New York launched similar protections and we’ve also seen similar liability protection efforts at the federal level, which Republican lawmakers are positioning as a requirement for any future stimulus packages.
- Harder and Faster Gallium Oxide Detectors: New research from North Carolina State University revealed that single-crystal gallium oxide radiation detectors could allow faster X-ray radiation monitoring (nearly in “real time”) and withstand higher levels of radiation than current X-ray detection materials. To validate this, the researchers made a radiation detector with a single-crystal sample of gallium oxide with electrodes attached on either side. They found that as X-ray radiation exposure increased, the gallium oxide current increased at the same pace.
- CV19 and VTE Imaging: A study from a Dutch team detailed the prevalence of venous thromboembolism (VTE) among COVID-19 patients, encouraging proactive diagnostic imaging efforts to identify deep vein thrombosis and pulmonary embolism and prevent VTE. The study of 198 hospitalized CV19 patients found that 16% of patients were diagnosed with VTE after 7 days and 42% were diagnosed with VTE after 21 days, while patients diagnosed with VTE had a much greater risk of death (2.4 HR).
- CV19 Space Tele-Ultrasound: Healthcare providers are using AdEchoTech ultrasound systems with European Space Agency-created robotics technology to remotely investigate medical conditions during the CV19 emergency. The system allows healthcare professionals with no ultrasound training to position patients in front of a robotic arm controlled by remote (and potentially quarantined / isolated) specialists.
- AI for Incidental CV19: Aidoc announced the FDA approval of its AI algorithm intended to detect and prioritize incidental COVID-19 findings in lung CT scans. Although coronavirus may seem like an unlikely incidental finding, Aidoc cited recent research that found 8-10% of patients undergoing CT scans for other conditions were discovered by radiologists to have COVID-19 and noted that the upcoming post-CV19 imaging surge could create a new wave of incidental findings.
- Braid’s $9M: Teleradiology startup Braid Health closed a $9m Seed and Series A round that it will use to grow its team, fund the development of its teleradiology platform, and develop its Braid Network of healthcare specialists. Braid Health intends to build a platform that combines AI-powered diagnostics and radiologists that will support providers ranging from retail pharmacies to health systems.
- CV19’s EU Volume Impact: A survey from IMV Medical Information Division detailed on AuntMinnieEurope.com found that 82% of European imaging centers experienced a major decline in non-CV19 imaging procedures (only 6% had “no decline”). Even when CV19 imaging is included, most modalities experienced volume declines with the greatest declines identified in mammography (-61.6%), MRI (-44.9%), fluoroscopy (-43.5%), and ultrasound (-40.2%), while mobile X-ray increased by a massive 51%.
- MEDNAX’s Imaging Down, Anesthesiology Out: MEDNAX’s Q1 briefing revealed that the COVID-19 emergency drove a 50% to 60% drop in April imaging volume, brought the mega specialist group to a $16.1m Q1 net loss, and prompted it to sell its anesthesiology division to North American Partners in Anesthesia (NAPA) for $50m (while keeping the division’s $110m AR balance). MEDNAX will use the funding from the anesthesiology acquisition to prop-up its existing businesses during the CV19 crisis.
The Resource Wire
– This is sponsored content.
- Hear Penn Medicine radiology professor Warren B. Gefter, MD describe how Nuance’s AI-driven technology enhances radiologist workflows and drives tangible patient outcomes.
- This Hitachi blog details the benefits of multi-parametric MRI (mpMRI) for prostate imaging, including insights on how its Oasis 1.2T High Field Open MR allows more diagnostic-quality exams with reduced metal susceptibility artifacts on the diffusion sequence compared to a 3T magnet.
- Join Healthcare Administrative Partners’ webinar, “Reentering the Post-COVID-19 Radiology Market,” on Wednesday, May 13 where we will discuss several factors radiology practices should take into consideration to ensure safe and successful reentry into the market.
- Learn how Riverain’s ClearRead CT Vessel Suppress provides a powerful and intuitive view for clinicians through the suppression of vascular structures.
- Watch an informational video from Jeff Hersh, M.D., Chief Medical Officer of GE Healthcare on using ultrasound in a health crisis.
- This Focused Ultrasound Foundation video details how focused ultrasound technology is advancing cancer immunotherapy.