“Talk to us so we can persuade you that this is not hype, but a game-changer for all humanity.”
Nanox Imaging founder, Ran Poliakine, inviting the major medical imaging players to consider adopting Nanox’s 5D X-ray source technology and help change the game for humanity.
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
Nanox’s Big Goal
Nanox Imaging, an Israeli/Japanese developer of a novel 5D X-ray source technology, announced a new investor and the launch of its global “1x1x1 Initiative,” both of which will play a key role in its long-awaited commercial rollout. This was a packed announcement revealing some massive goals and intriguing technology from a relatively unknown imaging hardware player.
- Everyone Gets Screened – Nanox’s 1x1x1 Initiative aims to screen every single person on the planet for cancer at least once each year by 2030. That’s a tall order given the limited healthcare infrastructure in many parts of the world, the overdiagnosis and workflow ramifications of screening the world’s entire population every 12 months, and the fact that many physicians would disagree with this goal (it’s also a 40x increase from today’s volume). However, this initiative does bring awareness to the fact that an estimated two-thirds of the world’s population does not have access to medical imaging, which directly supports Nanox’s product strategy.
- Small, Inexpensive, Fast – Nanox’s small size/low-cost value proposition would be quite relevant if the 1x1x1 Initiative (or any initiative to expand imaging access) somehow came to fruition. Nanox’s novel 5D X-ray source is based on nano MEMS technology, which creates a stable X-ray stream at low voltage, with far lower cooling requirements. This allows the Nanox X-ray technology to sell for a fraction of the cost of a CT (~$10k vs. $1 – $3m), scan up to 30x faster, and fit into a suitcase, while achieving what Nanox states is superior image quality.
- Fully Funded – After 15 years of development and $1 billion in startup funding, Nanox just gained new funding from Korea’s SK Telecom to support the new X-ray source’s commercialization, targeting 15,000 units deployed by 2022. SK Telecom really believes in Nanox, calling it “one of the most influential companies of our era,” with the potential to save millions of lives and trillions of dollars in cancer-associated costs.
- Partner Path – Nanox’s path to achieve these massive targets is through technology partnerships, calling on the major healthcare tech players (plus AI developers and radiology consortiums) to talk with Nanox so it can “persuade (them) that this is not hype, but a game-changer for all humanity.” Once “persuaded”, these players would utilize Nanox technology in their own branded products and through their own channels.
Skeptics will definitely have fun with the audacity of the 1x1x1 Initiative. Still, one look at Butterfly Network and the other handheld ultrasound players and it’s clear that expanding imaging access in developing regions is a solid play (for business, publicity, and healthcare ethics reasons), so it could also be argued that Nanox is on to something. Just because imaging systems are massive and expensive doesn’t mean they should remain that way forever.
AI Marketplace 101
Harvard Business Review and Nuance Healthcare teamed up to detail the role and benefits of radiology AI marketplaces, providing a basic introduction for HBR’s non-radiologist readers and bringing more attention to this nascent part of the AI ecosystem. Here are the big takeaways:
- The Problem – AI helps improve radiologist efficiency and patient care, but it’s still hard to put into everyday clinical use due to the abundant (and growing) list of single-function “narrow” AI solutions that radiologists have to seek out, evaluate, and implement.
- The Marketplace Solution – Marketplaces provide a centralized “last mile” platform to discover and distribute algorithms, create a feedback channel between users and developers, and support AI’s lifecycle requirements (developing, training, regulatory approval, deployment, and validation).
- The Partnership Play – AI Marketplaces’ feedback components may seem secondary to its distribution and lifecycle advantages, but this feedback is key for sharing/validating results and facilitates public-private partnerships to drive AI development (e.g. FDA’s Software Precertification Program and National Evaluation System for health Technology).
After nearly two years and amassing long list of AI partners, Nuance could be seen as a relative veteran in the young AI marketplace arena, and this HBR story is a great example of how Nuance has built its marketplace brand. Still, Nuance is not alone, as the above business and clinical motivations brought a wave of other AI platforms in recent years (Arterys, GE Healthcare Edison, Envoy AI, Incepto to name a few), revealing a growing industry consensus that marketplaces may be AI’s rightful “last mile” and highlighting how critical it is for these players to solidify their positioning as radiology’s go-to AI marketplace while the AI’s “last mile” is still taking shape.
NYT Gets Specific on the China Scandal
The Chinese medical imaging bribery scheme first revealed by Reuters a few weeks ago was brought to a much wider audience (with a lot more details) in a New York Times expose last week. Here’s what we learned:
- Degrees of Separation – Payments to public hospital officials were executed through local middlemen who often ordered devices through third party distributors or dealers (not OEMs)
- Scheme Structure – Middlemen inflated costs to fund bribes, avoided competition, and masked schemes through multiple distribution layers, while public bids were often held to maintain appearances even though deals were already decided
- Big Payouts – Highlighted bribes included $1m to win a $4m CT deal, a $900k bribe for another CT deal, and $400k to sell a $1.7m MRI (often using cash)
- Cases Closed – The Chinese government prosecuted at least 19 cases, with corrupt buyers sentenced to up to 15yrs and middlemen reps sentenced to up to 3yrs (OEMs not charged)
- This is Common – Although previous reports on the China and Brazil scandals generally focused more on GE, the NYT story positioned this as more of a brand-agnostic industry scandal (sometimes bribes even covered multiple brands) and “an unspoken industrial rule” in China
- OEMs Off the Hook so Far– Although OEMs were protected by the China scandals layered distribution structure, some argue that OEMs were well aware of these schemes, noting that both Chinese and U.S. laws technically hold manufacturers responsible for the actions of lower-level distributors
Although these Chinese cases are now complete, the U.S. federal authorities are still investigating these companies as part of a wider effort to enforce corruption in the global medical equipment market, suggesting that more stories like this are still coming.
It’s Difficult to Make Predictions, Especially About the Future
Wall Street Journal tech columnist Andy Kessler shared some well written reassurance that predictions of AI changing the working world as we know it are unlikely to come true. Kessler tracked through humanity’s long history of inaccurate tech and economic predictions (e.g. 60’s space utopia race, the Soviet nuclear race, 1999’s dot com bubble), and shared that:
“My experience is that people tend to overestimate the absurd, like Elon Musk’s dreams of building a hyperloop and colonizing Mars, and underestimate the mundane, like improvements in messaging and shopping. I’m usually bullish until dreams become hallucinations. Technology develops in S curves: Things start slow, go into hyperbolic growth, and then roll over. That’s why “the singularity”—self-improving, unrestrained artificial intelligence—probably won’t happen. Don’t extend the trend.
“The tempests of change blow hard. Reading the prevailing winds, we’re all about to become robot-replaced, drone-delivered-synthetic-meat-eating, augmented-reality-helmet-wearing, bitcoin-spending, fruit-flavored-vaping, neutered democratic socialists chirping “Comrade” and streaming “The Handmaid’s Tale” Season 10, “Dystopia’s Discontents,” on our watches while collecting universal basic income.”
It’s a fun read and a good reminder to keep calm and skeptic on when faced with cataclysmic forecasts. It’s also a useful reminder that fear of AI isn’t just a radiology thing.
- HeartFlow launched a nearly $100 million equity funding round, with $65 million already raised as of June 11th. This latest equity round would grow the FFRct leader’s total funding to over $576 million, following a $240 million Series E round in 2018. HeartFlow’s fundraising is particularly impressive given that most other AI leaders have raised between $30 million and $80 million so far, showing the potential that investors see in HeartFlow’s FFRct capabilities and the amount of capital that’s available to AI firms that truly standout.
- UK scientists are about to begin a “groundbreaking” clinical trial of a new 10-minute / £150 prostate MRI scanning method that scientists say could “transform” prostate cancer screening. The £5 million trial on 1,000 men (ages 55-77) launches after a change in national guidelines made MRI the initial prostate diagnosis step (vs. blood tests) and is intended to validate whether the new 10-minute method (a simplified version of a 30-minute scan for at-risk men) could be used in a national screening program.
- MRI scan speed is one thing, but a team from The Ottawa Hospital found that integrating the role of “director of prostate imaging” within geographic regions also helps improve prostate MRI performance and reporting. The team looked at prostate mpMRI images from four peripheral institutions (using five 1.5T MRIs) before and after the introduction of the new prostate imaging director role, achieving improvements in PI-RADSv2 compliance, T2-weighted spatial resolution compliance (up 40%), DWI spatial resolution compliance (20% to 100%), modified DWI spatial resolution compliance (60% to 100%), dynamic contrast-enhanced imaging compliance (60% to 100%), temporal resolution compliance (20% to 100%), and acquisition time compliance (60% to 100%).
- Konica Minolta Healthcare announced the launch of Konica Minolta Payment Services, a new in-house financing services program for its Exa Enterprise Imaging platform. Konica Minolta emphasized the program’s flexibility, as it can adapt to customers’ current operating model and cash flow needs (e.g. monthly capital payments), potentially allowing a more defined project scope of work and shortening purchasing times.
- There’s no doubt that radiologists are under pressure to work faster, resulting in differing opinions regarding diagnostic speed’s impact on accuracy (some say speed leads to greater accuracy, some disagree). A new paper published in the American Journal of Roentgenology might not have satisfied either side of that debate, finding that diagnostic speed is unlikely to influence accuracy and suggesting that most previous studies on this issue were based on insufficient or anecdotal evidence.
- IHS Markit laid out a useful roadmap on how ultrasound can realize the type of developing market growth that many have forecasted, suggesting that although ultrasound’s success in China serves as an attractive model to shoot for, certain things will have to happen for this success to be repeated in other emerging markets (specifically Latin America, the Middle East, Africa, and the rest of Asia Pacific). Specifically, most emerging markets need major government funded healthcare infrastructure initiatives (like Healthy China 2020) before the ultrasound segment can reach its potential and some (e.g. Latin America and the Middle East) will also have to overcome economic and political instability.
- Zebra Medical Vision announced the FDA clearance of its HealthICH product, an AI-based solution that identifies signs of intracranial hemorrhage (ICH) in head CT scans and alerts physicians if ICH is suspected. HealthICH is part of Zebra’s Multi-Modality AI Triage Solution, which provides alerts for intracranial hemorrhages (in head CTs) and pneumothorax (in chest X-rays), and integrates with the company’s “All in one” (AI1) radiology workflow solution. HealthICH is Zebra Med’s 3rd FDA cleared product, coming just a month after its pneumothorax solution.
- Huffpost revealed that U.S. citizens are increasingly resorting to GoFundMe campaigns to handle their medical bills, as 2017 included 250,000 medical fundraisers raising $650 million. This isn’t specifically an imaging story, but it certainly is a patient care story, it influences the way many in the U.S. view our healthcare system, and could be seen as a leading indicator of larger changes to how medical services are covered.
- Canon Medical Systems announced the FDA clearance of its long-awaited Advanced Intelligent Clear-IQ Engine (AiCE) CT image reconstruction technology, intended for use with its Aquilion CT scanners. AiCE leverages DCNN image reconstruction technology to reduce noise and enhance signal, allowing it to reconstruct images with improved spatial resolution and achieve 3-5-times faster speeds than traditional MBIR.
- Israeli researchers developed an AI algorithm, combining ML and DL, that’s able to predict breast cancer based on mammography images and health record data as accurately as radiologists, suggesting that it could be beneficial as a second reader. The study used 52,936 images and at least one year of pre-mammogram health records from 13,234 women, training the algorithm on mammograms and health records from 9,611 women to predict biopsy malignancy (identified 34 of 71 false-negatives) and differentiate normal from abnormal screening exams (0.91 AUC, 77.3% specificity, 87% sensitivity).
- iCAD announced the European launch of its ProFound AI for 2D Mammography solution, joining ProFound AI for DBT that launched in the U.S. and Europe in 2018, allowing European iCAD customers to support breast cancer workflow from both mammography technologies.
The Resource Wire
– This is sponsored content.
- By partnering with Medmo, imaging centers can keep their schedules full and their equipment busy. Here’s where to learn more and get started.
- This Qure.ai blog highlights how its qXR automated chest X-ray solution has helped scale tuberculosis screening to remote regions across the globe.
- In this Carestream video, an Orthopedic surgeon discusses how the OnSite Extremity CT has improved his business and the care he provides his patients.
- In this interview, the Cancer Research Institute leader, Jill O’Donnell-Tormey, PhD, discussed CRI’s partnership with the Focused Ultrasound Foundation, their vision for the combination of focused ultrasound and immunotherapy, and their achievements so far.
- Nuance’s updated PowerScribe One radiology reporting platform now provides access to AI-powered diagnostic and decision-support tools within radiologists’ familiar workflow, giving them tools to improve efficiency and throughput, increase accuracy and specificity, and ensure evidence-based follow-up.
- POCUS Systems is approved as a Veteran Owned Business with the US Government Office of Veterans Business Development, paving the way for partnerships with the federal healthcare delivery systems.