Imaging Coronavirus | AI’s Instantaneous Radiation Plans

“If you didn’t know about this outbreak, you’d read the scan and you would just say, ‘Okay, this patient has pneumonia’ . . .”

Thomas Jefferson University radiologist, Paras Lakhani, on how novel coronavirus scans might have been read before everyone was on the lookout for it.



The Imaging Wire is happy to announce that Healthcare Administrative Partners is now an Imaging Wire sponsor. Healthcare Administrative Partners does great work to empower radiology groups and we’re excited to share their message with you.

Every issue of The Imaging Wire is made possible by Healthcare Administrative Partners and the rest of our sponsors, so make sure to keep up with the great things they’re doing.


Imaging Wire Sponsors

  • Carestream – Focused on delivering innovation that is life changing – for patients, customers, employees, communities and other stakeholders
  • 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

The Imaging Wire

Imaging Coronavirus

The emergence and spread of novel coronavirus captured much of the world’s attention in recent weeks, as the number of infected people reached 10,000 (maybe much more) and the death toll surpassed 300. News of coronavirus has also highlighted the key role of medical imaging in understanding, diagnosing, and treating the disease. Here’s some of the big imaging angles in this very big story.

Image Observations – The Radiology journal published chest CT scans from a 33-year-old Chinese coronavirus patient, noting that the scans show white patches with ground glass opacity in the lower corner of her lungs that worsened in scans captured three days later. Observations of the images by a Thomas Jefferson University radiologist suggested that the scans could be mistaken for pneumonia, but the fact that the ground-glass patches extend to the edges of the patient’s lungs are unique and more similar to scans from SARS and MERS patients.

NEJM & CXR – Chest X-rays played a prominent role in a pair of NEJM reports, including one report that concluded that coronavirus can be transmitted between humans (not just from animals) and another that covered the first coronavirus case in the U.S. In both reports (a 65 year-old Vietnamese man and a 35-year old Washington state man), chest X-rays were used to support diagnosis (along with throat swabs) and measure the disease’s progression.

The Coronavirus Business – Coronavirus is also creating business and PR opportunities, as Siemens Healthineers announced a CT and ultrasound deal with the city of Wuhan (donating one of the ultrasounds), Canon donated a CT system to a hospital in Wuhan, Philips donated €2 million worth of devices (CTs, USs, other devices), and GE Healthcare revealed that demand for its R860 respirator system has increased by 20%.



AI’s Instantaneous Radiation Plans

A UT Southwestern Medical Center team developed four AI models that use imaging scans and other data to “instantly” generate radiation dosage plans (vs. a several days to a week turnaround), shortening the amount of time patients have to wait before starting radiation therapy.

Impact – Models like these could have life-saving and time-savings implications. Delaying radiation therapy by even a week increases the likelihood of some cancers recurring or spreading by 12% to 14%. Not to mention the fact that creating treatment plans “can be a time-consuming and tedious process” and these new models could “cut out much of the back and forth that happens between the doctor and the dosage planner.”

The First StudyThe UT team fed data from 70 prostate cancer patients into four deep-learning models, which produced 3D renderings of how best to distribute each patient’s radiation in five-hundredths of a second. The renderings accurately predicted the treatment plans developed by a medical team, achieving <2.8% and 4.2% respective mean and max dosage error rates.

The Second StudyThe team’s second study used an algorithm based on two of these dosage planning models (one that was fast, another that was accurate) that reduced the time required to recalculate dosages before each radiation session from 10+ minutes to one second.


Thank you all for supporting The Imaging Wire and Visage Imaging‘s referral program benefiting the New South Wales Rural Fire Service and WIRES Wildlife.

If you haven’t done so already, please take a moment to consider your colleagues who might find this newsletter valuable and let them know about it. Once again, we’re donating $2 to these Australian Bushfires charities for each new subscriber that you refer.

If you’re a subscriber, please consider using your personal referral link to introduce your industry friends to The Imaging Wire (You can find your personal link in your latest newsletter).

If you haven’t received a referral from one of your colleagues yet, just use this link to sign up.


The Wire

  • The results of the NELSON Dutch-Belgian lung cancer trial revealed that volume low-dose CT screening was indeed effective in reducing lung cancer mortality. The study (n = 15,792) randomly assigned participants to undergo LD-CT screening at different periods (baseline, 1, 3, 5.5, none). Ten-year follow-ups revealed lung cancer incidence was 5.58 cases per 1k person years in the screening group (vs. 4.91 in control group), while the screening group had much lower rates of lung cancer mortality (2.5 per 1k years vs. 3.3 per 1k years).
  • The ACR quickly acted on these NELSON trial results, calling on the U.S. federal government to adjust Medicare reimbursement rates to encourage nationwide low-dose CT lung cancer screening and leverage anti-smoking campaigns to expand public awareness of screening’s benefits.
  • Colorado’s St. Mary’s Medical will save its patients $5-$7 million this year by making big cuts to its imaging center’s MRI ($3k to $800), CT ($1.8k/$2k to $600), and ultrasound ($1.3k to $400) rates (mammography staying the same). The Grand Junction-based hospital made this decision to “give back” to its patients after achieving $12 million in improved cost efficiencies within other departments, but also apparently had (justifiable) business motivations for these cuts, noting that its new rates led to a “big increase” in imaging referrals and will help “keep competition out.”
  • RADPEER has racked up over 41 million peer reviews during its ~17 years, but an ACR survey (n = 300 practices) found that 31% of practices are still underreporting significant discrepancies due to concerns over the legal (56% concerned), emotional/relationship (20% to 52%), and professional (38%) consequences of reporting peers’ discrepancies. Another recent study found that integrating anonymous/randomized peer reviews into PACS workflows led to a fivefold increase in serious discrepancies reporting, suggesting that these changes could also help RADPEER overcome its users’ emotional barriers.
  • A team at Detroit’s Ascension St. John Hospital found that a 90-second magnetocardiography (MCG) scan can detect coronary artery stenosis (CAS) with similar accuracy as stress testing (ST) and coronary angiography (CA), potentially reducing the time it takes to spot patients who need care and those who are safe for discharge. The study scanned 101 patients who visited an ED with non-high-risk chest pain with MCG as well as ST and CA. Compared to ST and ST+CA, MCG had 95% and 33.3% sensitivity, 77.8% and 78.3% specificity, 13% and 13% positive predictive value, and 89.7% and 92.3% negative predictive value.
  • Imaging lobbying groups held a briefing on Capitol Hill to highlight the importance of amyloid PET imaging for diagnosis, research, and treatment of Alzheimer’s disease, while calling for the passage of the Medicare Diagnostic Radiopharmaceutical Payment Equity Act of 2019 (H.R. 3772). The groups criticized CMS’ treatment of radiopharmaceuticals as part of the packaged cost of a scan procedure, saying it disincentivizes radiopharmaceutical use in Medicare hospital outpatient settings, and would be addressed by the radiopharmaceutical act by increasing PET radiotracer reimbursements.
  • A study in the Journal of Nuclear Medicine identified new breast cancer biomarkers that can be identified using PET/MRI scans. The study scanned 141 women with F-FDG PET/MRI (100 malignant, 41 benign lesions), finding differences in the BPE, BPU, and FGT biomarkers between the women with malignant and benign lesions, and also discovering that the biomarkers are present in the women’s healthy breast (not just the breasts with tumors).
  • As we wrap up the first month of the Medicare AUC Mandate’s 2020 “educational and operations testing period,” Healthcare Administrative Partners outlined how radiology practices can prepare for 2021’s higher-stakes Medicare AUC period. The firm recommended that radiology groups use the 2020 testing year to gather data on their ordering practices, provide targeted marketing toward high-volume practices, and make sure referring practices are using a CMS-qualified Clinical Decision Support Mechanism (CDSM). Once CMS begins collecting AUC data in 2021, Healthcare Administrative Partners suggests that radiology practices get serious about adherence, as initial data collected by CMS will be used to identify “outlier ordering professionals” who would face additional prior authorization requirements in the future.
  • A Definitive Healthcare survey of 207 imaging leaders and professionals found that one-third of U.S. facilities are using AI to aid imaging studies and another third plan on adopting AI within the next two years. This may seem high for a pretty new technology, but there’s a very established reason for these adoption rates, as 93% of the current adopters use AI for CAD (vs. 27% for process/workflow improvement, 16% for tech/equipment monitoring, and 13% to guide consultations and suggest care).
  • Siemens Healthineers and Hamilton Health Sciences announced a $270 million value partnership that will provide the Ontario, Canada-based health system with Siemens Healthineers imaging systems and related services over the next 15 years (and potentially 25 years). The agreement includes the procurement, ongoing replacement, and maintenance of approximately 500 imaging systems.
  • New analysis in JACR details how an MGH team is developing new AI-based CT image reconstruction algorithms that could lead to “the Holy Grail of high-image quality at low radiation dose,” overcoming the image quality concerns that surround current image reconstruction algorithms.
  • Agfa-Gevaert and Dedalus signed their long anticipated share purchase agreement, selling Agfa HealthCare’s IT business to Dedalus for €975 million (well above the division’s $288m annual revenue). The deal also includes Agfa’s Imaging IT business in Germany, Austria, Switzerland, France, and Brazil (~17% of its imaging IT revenue) because the two divisions are so tightly integrated in these regions.

The Resource Wire

– This is sponsored content.

  • The 2020 CPT code updates brought a number of changes to diagnostic radiology, interventional radiology, and nuclear medicine. This Healthcare Administrative Partners blog details the 2020 MPFS CPT code changes and what they mean to your practice.
  • This MEDNAX blog details how Qure.ai worked with the imaging center giant to validate its qER algorithm against radiologist findings on more than 300,000 head CT scans, achieving an AUC of 0.96.
  • 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.
  • 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 Carestream blog highlights the top reasons healthcare providers should make the shift to digital.

2 Responses to “Imaging Coronavirus | AI’s Instantaneous Radiation Plans”

Leave a Reply

Basic HTML is allowed. Your email address will not be published.

Subscribe to this comment feed via RSS

Get every issue of The Imaging Wire

Delivered right to your inbox. We will never sell or share your information.