“The lowest branches of the diagnostic tree are bare, with every low hanging fruit having been picked, often multiple times.”
Dr. Hugh Harvey and Dr. Shah Islam on why it’s too late to launch another diagnostic AI startup.
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
- Medmo – Helping underinsured Americans save on medical scans by connecting them to imaging providers with unfilled schedule time
- Nuance – AI and cloud-powered technology solutions to help radiologists stay focused, move quickly, and work smarter
- POCUS Systems – A new Point of Care Ultrasound startup, combining a team of POCUS veterans with next-generation genuine AI technology to disrupt the industry
- Qure.ai – Making healthcare more accessible by applying deep learning to radiology imaging
The Imaging Wire
Hardian’s AI Roundup
If you follow Dr. Hugh Harvey on Twitter, you may have noticed that he has some strong opinions on the business of radiology AI and doesn’t hold back when he sees something that falls outside of best practice. Dr. Harvey and his Hardian Health colleague Dr. Shah Islam brought that same spirit to their latest blog reviewing the RSNA AI Showcase and may have hurt some feelings in the process.
- Where’s the AI Showcase? – As mentioned in our own RSNA wrap-up last week, the blog criticized RSNA’s decision to “segregate” the AI exhibitors, noting how the AI Showcase reduced foot traffic and contradicted this year’s theme of AI integration.
- Don’t Launch an AI Company – This may not be a long term problem for RSNA, as Dr. Harvey and Dr. Islam argued that we’ve reached the diagnostic AI startup saturation point and advised that the best way to build an AI company is to not build one at all. Meanwhile, they believe “the well-published (and non-VC funded startups) stand the best chance of survival.”
- There Were Still Highlights – The blog listed through a handful of AI startups who deserved praise (Qure.ai, Viz.ai, Aidence, Lunit, Icometrix, Aidoc, MaxQ AI, Quibim, Riverain and Zebra), while also pointing to the AI application areas that aren’t yet saturated (automated ultrasound diagnostics, biomarker-driven insights, AI-enhanced image recognition/reconstruction), and highlighting a much-needed focus on structured reporting.
- Triage and Evidence Critique – Although many AI solutions tout worklist prioritization as a core part of their value proposition, the article argues that radiologists’ desire for control and reliance on read-based compensation will stand in the way of adoption (at least in the U.S.). Conversely, Harvey and Islam argue that diagnostic AI’s greatest benefit will come from improving accuracy of non-time-critical pathologies “that have a significant downstream cost when not detected on initial imaging” but those applications will still require evidence from prospective studies.
- Shaky AI Financials – Hardian Health reminded readers of the recent rumors that AI startups are increasingly hitting financial rough patches, reiterating that AI startups still largely rely on VC funding and although some are gaining traction “the uptake is painfully slow.” After all, this is medicine where “procurement cycles take eons.”
- AI Marketplaces – The blog ran through the list of AI marketplaces (a list that got even longer at RSNA), suggesting that they’ll face an uphill battle versus the OEMs and PACS providers’ major integration advantages. With that, Harvey and Islam suggesting that the pure play AI platforms should focus on “white-labelling their software to the big fish.”
- Rise of Remote Reporting AI – The Hardian team forecasted that AI adoption will fall dramatically after adoption from large hospitals is complete, but the delayed adoption from mid-sized hospitals will actually lead to growth in AI-aided teleradiology. This is where they see the greatest AI growth opportunities and it makes sense given teleradiology’s high 24/7 volumes, focus on human and machine efficiency, and strong IT capabilities.
Siemens’ New CTs & AI
Siemens Healthineers’ RSNA launches largely focused on its CT lineup and AI portfolio, while the company also introduced a new rad tech staffing service, and debuted its previously announced ultrasound and angio systems.
- SOMATOM On.site – Siemens made a pretty notable addition to its CT lineup with the SOMATOM On.site mobile head CT (FDA pending), immediately making Siemens the most prominent OEM in the head CT segment. The SOMATOM On.site is intended to provide bedside scans in the ICU, significantly reducing the time (~1hr) and staff (up to 5 people) traditionally required to transport patients to a CT scanner.
- SOMATOM X.cite – Siemens displayed its recently FDA cleared SOMATOM X.cite premium single-source CT scanner (82cm bore, Vectron tube, Patient Observation Camera), highlighted by the company’s new myExam Companion intelligent UI concept.
- New AI – Siemens expanded its AI portfolio, introducing the AI-Rad Companion for Morphometry Analysis (provides result tables on volumetric changes in the brain) and AI-Rad Companion Prostate MR for Biopsy Support (segments prostate, facilitates MRI/US targeted biopsy). The new solutions add to Siemens’ AI-Rad Companion portfolio that launched at RSNA 2018.
- CrewPlace – Siemens debuted its CrewPlace cloud-based technologist workforce platform, used by healthcare providers to manage their radiologic technologist workforce and (if needed) access Siemens’ pool of outside rad techs to offset staffing shortages or expand services beyond their teams’ current skillset.
- Acuson Redwood – The Acuson Redwood multi-department ultrasound (radiology, cardiology, and ob/gyn) made its RSNA debut after gaining FDA and CE clearance in October.
- Artis icono – Siemens showed its recently FDA-cleared Artis icono angiography systems, including the “Artis icono” biplane system (for neuroradiology and abdominal procedures) and “ARTIS icono floor” single-plane floor-mounted system (for vascular, interventional cardiology, surgical, and oncology procedures).
- A new study in The Journal of Nuclear Cardiology (n = 4,619 PET patients, 4,619 stress echo patients) found that stress echocardiography deserves its role as the go-to method for evaluating suspected coronary artery disease (used in 96.8%of all cases), and not cardiac PET (used in 3.2% of cases). The study found that cardiac PET did not reduce utilization or improve outcomes enough to offset its higher costs ($2,358 vs. $1,774 per month).
- An activist investor is pushing for a full or partial sale of Mednax, with reports suggesting that a proxy battle may take place if the company’s board doesn’t agree to the plan. This may not be surprising to some, given Mednax’s financial direction (shares down 25% YTD, net income down 4-straight years, falling income) and the company’s recent move to halt radiology M&A.
- A VA inspector general report revealed that VA employees canceled and delayed tens of thousands imaging orders, attributing the issues to backlogs, breakdowns, and mismanagement. The report confirms previous allegations that the VA cancelled 250,000 radiology orders between 2016 and 2018 without performing the appropriate clinical review. VA officials pledged to address this issue by July 2020.
- Meanwhile, the VA announced the launch of its National Artificial Intelligence Institute (NAII), focused on developing AI R&D capabilities, initially targeting deep learning, explainable AI, privacy-preserving AI, and AI for multi-scale time series.
- RSNA wrapped up with a Johns Hopkins Hospital study (n=229,464 patients) revealing that although the vast majority of patients with a new diagnosis of carotid artery stenosis receive ultrasound scans (97% in 2011 to 94% in 2016), the use of CT angiography has increased significantly (CTA, 1.6% in 2011 to 4.7% in 2016). The team expressed concern over CTA’s growth, given its higher cost and dosage and the fact that that carotid artery duplex ultrasonography (DUS) is the recommended first-line imaging modality for carotid artery stenosis diagnosis.
- Synaptive Medical completed a $25 million equity round and converted roughly $50 million of its debt to equity, revealing plans to use its new funding to expand its cranial and spinal imaging product lineup (surgical planning, research). The deal also resulted in a manufacturing agreement with investor, Linamar.
- United Imaging’s RSNA booth brought the debut of its new uMR Omega MR system, which is highlighted by its ultra-wide 75-cm bore, 680 lbs table weight capacity, and wide 60x60x50cm imaging FOV. The uMR launch announcement focused on how its large bore supports the imaging experience and improves access for overweight and anxious patients.
- The FDA scheduled a two-day public workshop in February to explore the benefits and risks of AI’s role in radiology, with the primary goal of helping the FDA effectively validate AI algorithms.
- ViewRay announced collaborations with Elekta and Medtronic that will help advance its MRIdian MR-guided radiation therapy system/technology and raise at least $75 million in new funding. The Elekta collaboration will focus on growing the MR-guided radiation therapy market and give Elekta an up to 9.9% stake in ViewRay, while the Medtronic deal will explore the benefits of MRIdian MR-guided radiation therapy and give Medtronic a minority interest in ViewRay.
- After months of public and private lobbying, it looks like congress is getting close to a surprise medical billing deal that will favor patients and insurers, at the expense of providers. The latest proposal would require insurers to pay at least the median in-network market rate and allow arbitration for bills above $750, while stopping providers from billing out-of-network patients directly. This isn’t good news for providers who commonly bill out-of-network patients (like radiologists) and fear the new median in-network rate would limit their negotiation leverage with insurers.
- Samsung’s RSNA booth brought the introduction of the new RS85 Prestige ultrasound (not yet in the U.S.), which is based on the FDA-cleared RS85, and features improved image quality/processing and new clinical applications to support viewing blood flow and attenuated regions behind bony structures. NeuroLogica also debuted the DORO Skull Clamp for the OmniTom Mobile head and neck CT, used to improve patient positioning during surgical intervention.
- The Department of Defense (DOD) announced that it will cover DBT as primary screening option for servicewomen covered by Tricare starting in January, shortly after a pair of U.S. senators filed an act to require the coverage. Since the DOD adheres to recommendations from the United States Preventive Services Task Force (USPSTF), DBT coverage is only approved on a provisional basis that can be extended for up to five years.
- Agfa HealthCare launched its DR 100s mobile DR system at RSNA, touting its slim form factor (22.8” wide), convenience features (power cord positioning, storage), integrated detector battery charger and security locks, 10” tube head, and its use of Agfa’s prioritized MUSICA Acquisition Workstation.
The Resource Wire
- If you weren’t already following along, Qure.ai, RAD-AID, and Saurabh Jha recently had an excellent tweetchat about the use of imaging AI to fight TB and other diseases in developing regions.
- This Carestream Special Report details how providers can get the greatest ROI from their X-ray technology as radiography demands increase and budgets head the other direction.
- This Nuance video shares how Orlando’s Arnold Palmer Children’s Hospital used PowerShare to #ditchthedisk and improve patient care.
- The Focused Ultrasound Foundation’s 2019 State of the Field Report details FUSF’s initiatives and achievements and the state of the focused ultrasound market.
- Yale University research reveals that the average patient drives past six lower-cost providers on the way to an imaging procedure, due in large part to patients’ and physicians’ limited cost consciousness. Medmo helps address this issue by letting patients enter what they can afford for their scan, then booking them at a nearby imaging center willing to accept that rate.
- POCUS Systems’ forthcoming ultrasounds will combine ease of use, durability, and reliability, allowing clinicians to focus on their patients.