Another Shot at ABR | More DBT Evidence | AI is Different

“Thus, MOC is nothing more than a device to force radiologists to pay tens of millions of dollars in MOC-related fees for a redundant, worthless, and superfluous CPD product.”

One of the many burns included in Dr. Sadhish K. Siva’s legal argument against the American Board of Radiology’s MOC process.

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

Another Shot at the ABR

The Tennessee radiologist who filed an unsuccessful lawsuit against the American Board of Radiology (ABR) is giving his case another shot.

  • Initial Complaint – Nearly a year ago, Dr. Sadhish K. Siva launched a lawsuit against the ABR, alleging that the organization uses its monopoly over the initial certification process (IC) to create ongoing control over the maintenance of certification process (MOC).
  • Case Dismissed – The judge who dismissed the case saw the ABR’s IC and MOCs as two parts of one product, ABR certification, so therefore it wasn’t illegal that they are tied together.
  • Amended Complaint – Dr. Siva filed an amended complaint that positions the ABR’s MOC program as a “continuing professional development (CPD)” product (in other words a CME product). By redefining the ABR’s MOC as a CPD/CME product, and by pointing out that other CPD/CME groups don’t sell initial certification, Dr. Silva is trying to support his core argument that the ABR’s IC and MOC are separate products that are illegally tied together. Now the judge just has to agree.

The new amended complaint is quite detailed and this coverage from Ben White is really good, so those who want to dive deeper have some solid options.

More DBT Evidence

A new Brigham and Women’s Hospital-led study added more evidence supporting DBT’s advantages over full-field digital mammography (FFDM), finding that DBT achieved far greater cancer detection rates and Positive Predictive Values than FFDM or hybrid DBT+FFDM screening.

  • The Study – The retrospective study reviewed 179,028 outpatient mammography screenings from 2011 through 2017. Of those screenings, 41,818 (23.3%) were FFDM-only, 83,125 (46.4%) were hybrid, and 54,084 (30.2%) were DBT-only.
  • The Results – DBT-only exams had by far the greatest cancer detection rate per 1,000 exams (6 vs. 2.6 w/ FFDM, 4.9 w/ hybrid) as well as the highest PPV1 (5.6% vs. 2.5% w/ FFDM, 4.6% w/ hybrid). The three methods had very similar recall rates (10.4% FFDM, 10.6% hybrid, 10.8% DBT), which doesn’t support previous findings that DBT results in fewer recalls, but certainly supports DBT’s role as a screening option.
  • Significance – Studies like this help justify DBT’s rapid adoption, which some argue was mainly achieved through marketing and lobbying, despite a lack of clinical evidence. The study also provides a unique view into the performance of DBT-only environments, noting that many FFDM vs. DBT studies were observational and looked at mixed FFDM/DBT environments that could be prone to selection bias.

AI is Different

The AI business model is not the same as the software business model. That’s the gist of this recent post from Silicon Valley VC giant Andreessen Horowitz, which has plenty of experience with both types of companies, and provides some interesting observations and suggestions about AI’s differences:

  • AI is Different – Although AI startups and their investors may hope the AI business will come with the upsides of traditional software (high margins, scalability), Andreessen Horowitz is “not so sure” about that. In fact, the firm argues that AI companies have many trademarks of traditional service companies and are more likely to display a mix of the two business models.
  • What’s So Different About AI? – Compared to traditional software companies, Andreessen Horowitz finds that AI firms have lower gross margins due to high cloud infrastructure and human support costs (50-60% vs. SaaS’ 60-80%+ benchmark), scaling challenges due to laborious “edge cases” (e.g. needy clients, long implementations), and “weaker defensive moats” (aka less guaranteed competitive advantages) due to commoditization of AI models and challenges with data networks.
  • AH’s Advice – Different isn’t necessarily worse, as long as you realize the AI business model actually is different and you “own the challenges and combine the best of both services and software.” Here are some steps to do that: 1. Eliminate model complexity and have as many customers as possible use the same models (easier, faster, less infrastructure costs); 2. Go narrow when choosing problems to address (reduces threat of edge cases, improves scalability); 3. Plan for high variable costs and lower gross margins, and manage business accordingly; 4. Embrace services as an add-on opportunity (an AI+service hybrid model); 5. AI is still young, so plan for your tech stack to change; and 6. Build defensibility “the old-fashioned way” (good products with a good strategy).

Share The Wire

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.

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

  • Jubilant Wins Again: Jubilant DraxImage announced another win in its ongoing patent battle against Bracco, as the U.S. Patent Office invalidated two of Bracco’s Rubidium-Infusion-System patents, thus allowing Jubilant to continue to sell its RUBY-FILL Generator and RUBY Rubidium Elution System. Bracco initially alleged that the RUBY-FILL Generator and RUBY Rubidium Elution System infringed on the two patents, prompting Jubilant to (now successfully) challenge the patents’ validity.
  • Rad Mentors: A new study out of Canada (n = 152 residents, 20 in radiology, surveys scored 1-5) suggests that the radiologist career path’s recent fall in popularity is in part due to concerns over technology outsourcing (2.6 out of 5), its lack of patient contact (3.9), and it’s dark work environment (3.6). Although these factors aren’t very controllable, mentorship during med school may help guide more to students towards radiology, as residents who chose radiology reported positive radiologist mentorships and less concern about the threat of technology.
  • CEM, Breast MRI Alternative: Austrian researchers provided new evidence supporting low-dose contrast‐enhanced mammography’s (L-CEM) role as an alternative to contrast‐enhanced breast MRI (CE-MRI) that could help reduce false‐positive biopsies. The study scanned 80 women who had suspicious initial breast imaging findings (93 lesions) with CE‐MRI and L-CEM. Readings from three blinded radiologists found that L-CEM could achieve greater specificity (46.9–96.9% vs. 37.5–53.1%) and PPV (76.4–97.6% vs. 73.3–77.3%) than CE-MRI, with “comparable” sensitivity (65.6–90.2% vs. 83.6–93.4%) and NPV (59.6–71.4% vs. 63.0–76.5%). However, CE‐MRI’s detection rate was still much higher than L-CEM (92.5–94.6% vs. 79.6–91.4%).
  • Mach7 & Altamont: Mach7 will use Altamont Software’s CaptureWare solution to help Mach7’s Enterprise Imaging Platform (EIP) ingest more DICOM and/or non-DICOM data. The announcement particularly highlighted how CaptureWare allows Mach7 to support scanned paper documents, CD import and burning, and printing to DICOM.
  • AI Inconsistencies: The AI research pragmatists can bolster their arguments with a recent study out of China that detailed the drop in deep learning model performance when used with external data. The study used six mammogram classification models (3 already published, 3 developed for the study) and exposed them to four new mammogram datasets, finding that AUCs for the three published models (0.88-0.95 during validation) and the three study-developed models (0.71-0.79 during validation) declined significantly when used with the new external data sets (0.44-0.65).
  • PE M&A: A new JAMA article explored private equity’s growing role in physician practice acquisitions, noting that PE-backed acquisitions increased from 59 in 2013 to 136 in 2016 (355 4yr total), with anesthesiology (19%), multi-specialty (19%), and emergency (12%) practices experiencing the most acquisitions (radiology was ~4%). The paper warned of PE firms’ role in practice consolidation, calling for research into the impact of these acquisitions and whether PE firms’ financial goals conflict with proper practice management.
  • Thyroid Shielding Fake News: News fact-checking service, AFP Fact Check, addressed a number of viral social media posts that suggest screening mammograms are driving an increase in thyroid cancer and calling for the adoption of thyroid shielding during breast cancer screening. This theory apparently dates back to a 2011 Dr. Oz episode, although both Dr. Oz’s site and numerous experts clarify that mammography does not meaningfully increase women’s risk of thyroid cancer (thyroid shielding is still recommended for dental X-rays, though).
  • Konica Minolta & RegenLab: Konica Minolta Healthcare will distribute RegenLab’s regenerative medicine portfolio of bio-injectable products, expanding Konica Minolta’s UGPro Solution program that’s intended to combine its ultrasound systems with US-guided therapies and hands-on education. RegenLab’s bio-injectable product line includes preparation devices for Platelet Rich Plasma (PRP) and other blood cell therapies that are delivered with ultrasound guidance.
  • Coronavirus AI: Chinese AI developer YITU Healthcare released an imaging AI algorithm intended to support the detection of coronavirus/COVID-19. YITU’s new Intelligent Evaluation System of Chest CT supports the diagnosis and quantitative evaluation of CT images, providing insights into severity, treatment efficacy, and prognosis. The new YITU Healthcare algorithm joins another new AI product from Infervision intended to help detect and monitor Coronavirus.
  • Cigna’s Imaging Limits: Cigna will soon launch a new advanced imaging site of care policy that limits the circumstances that onsite MRI and CT imaging would be approved to: patients under 10yrs, obstetrical observations, perinatology services, transplantation services, patients with a known contrast allergy, and when there are no offsite alternatives or offsite imaging would adversely impact care. Anything else would be considered “not medically necessary” and would only be covered at a standalone imaging center.
  • Stop Means Stop: A Massachusetts appeals court ruled that a patient’s request to “stop” an imaging procedure is sufficient communication to withdraw consent. This isn’t good news for Brigham and Women’s Hospital, which has been involved in a legal dispute with the family of a now-deceased woman, who allege that BWH technologists completed an X-ray exam despite the woman’s pleas for them to stop. The court previously ruled in favor of BWH in a summary judgement, but the new ruling allows for a battery and emotional distress lawsuit.
  • BBV & HeartVista: Radiology AI-focused fund Bold Brain Ventures (BBV) announced its investment in AI-assisted cardiac MRI solution company, HeartVista. The funding comes about four months after HeartVista’s One Click Cardiac MRI Package software (guides MR image acquisition) gained FDA approval and about 16 months after BBV’s launch.

The Resource Wire

– This is sponsored content

  • 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.
  • The pressures driving radiology practice consolidation are significant, but there’s a strong case for staying independent. Healthcare Administrative Partners explains how independent practices can thrive by building relationships, relying on outside support, and leveraging capital.
  • ClearRead CT from Riverain Technologies is the first FDA-cleared system for the automatic detection of all lung nodule types, allowing radiologists to reduce search and reporting time and improve nodule detection rates.

You might also like

You might also like..

Select All

You're signed up!

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

-- The Imaging Wire team

You're all set!