IBM Watson Health’s stumble and possible fall

July 21, 2021 | by

This Editor hadn’t thought about or seen news about IBM Watson Health in over a year…and likely, neither did you. Granted, our minds have been Otherwise Engaged, but for the company that was supposed to dominate AI and health analytics, it’s notable that TTA’s last two articles mentioning Watson Health was 25 April 2019, on a report that its Drug Discovery unit was being cut back as the latest in a series of executive cutbacks and lawsuits (MD Anderson on a failed oncology initiative), and 14 Feb 2020 on 3M’s lawsuit on unauthorized use of their software.

The New York Times in an investigative piece (may be paywalled or require signup for limited access), brings us up to date on what is happening at IBM Watson, and it’s not bright for Watson Health. IBM, like so many other companies, badly underestimated the sheer screaming complexity of health data. Their executives believed they could translate the big win on the “Jeopardy!” game show in 2011, based on brute computing power, into mastery of healthcare data and translation into massive predictive models. The CEO at the time called it their ‘moon shot’. Big thinkers such as Clayton Christensen chimed in. IBM managers sang its praises to all in healthcare who would listen. This Editor, on a gig at a major health plan in NJ that was ‘thinking big’ at the time and used IBM consultants extensively, in 2012 was able to bring in speakers from Watson for an internal meeting.

But we haven’t been on the moon since 1972 (though probes have visited Mars). Since the big push in 2011-12, it’s been one stumble after another. According to the Times:

  • The bar was set much too high with oncology. Watson researchers knew early on in their research at the University of North Carolina School of Medicine that their genetic data was filled with gaps, complexity, and messiness. The experience was similar with Memorial Sloan-Kettering Cancer Center. The products growing out of the UNC and MSKCC research, Watson for Genomics and Watson for Oncology, were discontinued last year. These were in addition to the MD Anderson Cancer Center initiative, Oncology Expert Advisor for treatment recommendations, that was kicked to the curb [TTA 22 Feb 2017] after $62 million spent. At the same time, IBM’s CEO was proudly announcing at HIMSS17 that they were betting the company on multiple new initiatives. 
  • Watson Health, formed in 2015, bought leading data analytics companies and then didn’t know what to do with them. TTA noted in August 2018 that Phytel, Explorys, and Truven Health Analytics were acquired as market leaders with significant books of business–and then shrank after being ‘bluewashed’. HISTalk, in its review of the Times article, noted that along with Merge Healthcare, IBM spent $4 billion for these companies. IBM’s difficulties in crunching real doctor and physical data were well known in 2018 with revealing articles in IEEE Spectrum and Der Spiegel. 

Six years later, Watson Health has been drastically pared back and reportedly is up for sale. Smaller, nimbler companies have taken over cloud computing and data analytics with AI and machine learning solutions that broke problems down into manageable chunks and business niches.

What’s recoverable from Watson? Basic, crunchy AI. Watson does natural language processing very well, as well as or better than Amazon, Google, and Microsoft. Watson Assistant is used by payers like Anthem to automate customer inquiries. Hardly a moonshot or even clinical decision support. For business, Watson applications automate basic tasks in ‘dishwashing’ areas such as accounting, payments, technology operations, marketing, and customer service. The bottom line is not good for IBM; both areas bring in a reported $1 billion per year but Watson continues to lose money. 

Lightning news roundup: AI for health systems Olive scores $400M, VA’s sticking with Cerner EHR, Black+Decker gets into the PERS game

July 2, 2021 | by

As here in the US we are winding up for our Independence Day holiday (apologies to King George III)….

Olive, a healthcare automation company for healthcare organizations, scored a venture round of $400 million from Vista Equity Partners. To date, it’s raised $856 million through a Series G plus this round and is now valued at $4 billion according to the company release. Olive’s value proposition is automating via AI routine processes and workflows, such as benefit verification discovery, prior authorizations, and billing/payments for health systems. About 900 US hospitals have adopted Olive’s systems. Mobihealthnews.

Breaking: The US Department of Veterans Affairs will be staying with Cerner Millenium for their EHR modernization from VistA. This follows a 12-week review of the implementation following failures within the $16 billion program itemized by the Government Accountability Office (GAO) in February [TTA 19 Feb]. Secretary Denis McDonough is scheduling two further review weeks to determine additional changes to the program. The intent is to build a cloud-based system fully interoperable with the Department of Defense’s Military Health System (MHS) also built with Cerner. FedScoop, Healthcare IT News

And in the What Are They Drinking in Marketing? I want some of that, stat! department…

Black + Decker is now becoming a PERS provider with the introduction of Black+Decker Health and the goVia line of mobile and home-based PERS with optional fall extension and call center monitoring through Medical Guardian . The devices are a fairly predictable line of cellular-connected (Verizon, AT+T) with a ‘classic’ home landline unit. The units are being sold through Amazon. B+D release

From a marketing perspective, the Black+Decker name, identified for decades with home and power tools, on a PERS line is also a classic–a classic mistaken line extension like Cadbury mashed potatoes or Colgate frozen entrees. Buy a PERS, get a drill? Relevance and fit to a older, female-skewing group?  It surely looks like their parent Stanley, which is a leading company in institutional alarm and location services. offloaded this legacy business to them. (Judging from the website, someone’s in a rush as some pages still have ‘greek’ copy under headings.) Hat tip to a Reader who wishes to remain anonymous.

Mayo Clinic creates AI-powered clinical decision/diagnostics support platform, two digital health portfolio companies

April 23, 2021 | by

“Changing the nature of healthcare from episodic to continuous”. Mayo Clinic announced the launch of the Remote Diagnostics and Management Platform (RDMP) that will connect data to artificial intelligence (AI) algorithms and create a ‘next generation’ of clinical support tools, diagnostics, and care protocols for faster diagnostics and more continuous care. According to Mayo Clinic Platform president John Halamka, MD, “clinicians will have access to best-in-class algorithms and care protocols and will be able to serve more patients effectively in remote care settings.” Patients will be able to access information to take better control of their health and make more informed decisions.

Mayo Clinic, with partners, is also organizing two portfolio companies to support RDMP:

  • Anumana, Inc. With nference, a synthesizer of biomedical data, Anumana will bring to market digital sensor diagnostics to decipher electrocardiograms (ECGs). The objective is to more effectively spot heart disease at the pre-symptomatic stage, enabling early treatment that saves patients and costs. Their first project will be to develop neural network algorithms based on billions of relevant pieces of heart health data contained in Mayo’s Clinical Data Analytics Platform, including millions of raw ECG signals. nference with Mayo in the past year has released COVID-19 molecular research based on Mayo data. Anumana completed a Series A of $25.7 million funded by the partner companies plus Matrix Capital Management, Matrix Partners, and NTTVC.  nference release.
  • Lucem Health Inc. With Commure, a General Catalyst portfolio technology company that accelerates healthcare software development, Lucem will develop the platform for connecting remote patient telemetry devices with AI-enabled algorithms. Lucem is kicking off with a jointly funded $6 million Series A. 

We noted back in 2019 Dr. Halamka’s move to Mayo to head up a machine learning/AI initiative which took a while (during a pandemic year) but is moving quickly. The Mayo release includes a YouTube video of Drs. Halamka and Friedman explaining Anumana’s objectives in early diagnosis reading ‘those invisible signals’ well ahead of an event, especially needed with heart disease as the first symptom may be devastating or deadly. Hat tip to HISTalk, which also amusingly notes Dr. Halamka’s sartorial changes.

 

AI-powered contact tracing as part of an ‘application ecosystem’ for COVID-19 information and vaccination

December 18, 2020 | by

Following up with Avaya UK on October’s Perspectives on COVID-19 contact tracing and the use of AI to automate virtual agents for initial contact, they have released details on their contact tracing application that integrates with current Avaya software or into a company’s call center system. Based on their materials, it automates the initial contact with the individual using natural language text to a smartphone or tablet messaging app, web chat, or email. AI in the contact tracing app helps to screen the response and directs it to the correct agent. Augmentation tools provide real-time prompts and suggestions during a live call with the individual. Notifications can also be automated and also individual follow up can be made via text message. Additional features are detailed on their web page and in the contact tracing overview (PDF). Having heard horror stories from friends who have been subject to contact tracing and follow up apps in the wake of COVID-19 contacts and diagnoses, a great annoyance was daily live phone calls with agents repeatedly asking the same information and making the same assistance offers. Text messages would have been far more acceptable and directive.

Contact tracing is a part of their OneCloud Communications Platform as a Service (CPaaS) which enables organizations to design their own applications and workflows with a platform that supports SMS, MMS, voice, messaging, transcriptions, and digital channels. With vaccination now front and center, for provider organizations, OneCloud can be used to build systems for COVID-19 vaccination information access, recruiting staff, and administering the process. Additional details are in their OneCloud CPaaS overview.

This week, OneCloud for healthcare was awarded Frost & Sullivan’s Competitive Strategy Leadership Award. Release.

Hat tip to Mary Burtt of AxiCom UK

Early detection of Parkinson’s via AI (and a surprising medium); Ed Marx on the digital transformation (or not) of health systems and COVID treatment at home

November 20, 2020 | by

Somewhat off our normal beat….but of interest.

Ardigen and The BioCollective are collaborating on early detection research for Parkinson’s Disease, based on a microbiome-based biomarker. Ardigen has developed an Artificial Intelligence (AI) Microbiome Translational Platform. The BioCollective has a bank of metagenomic and patient metadata generated from an unexpected source: Parkinson’s patients’ stool samples. Release

The BioCollective is headed by Martha Carlin, who came from well outside of healthcare and pulled together a research group to address her husband’s diagnosis. A visit to this website is worth an examination on how these samples are collected for microbiome extraction. An interesting twist is the marketing of a probiotic mix developed using their BioFlux metabolic model for ‘gut health’.

Ed Marx, the former CIO of the Cleveland Clinic, has written a new book, ‘Healthcare Digital Transformation: How Consumerism, Technology, and the Pandemic are Accelerating the Future’. It’s billed as a wake-up call for healthcare systems and hospitals under challenge by Big Retail, Big Pharma, and Big Tech. This Editor met Mr. Marx when he premiered his entertaining memoir, ‘Extraordinary Tales from a Rather Ordinary Guy’, a few years ago. On treatment for COVID patients, except for the very sickest, he advocates it being done from home. From the release: “When the pandemic hit, a lot of progressive organizations would send most of their Covid patients home with monitoring equipment hooked up to phones unless they needed a ventilator. It’s a lot cheaper than staying in the hospital.”

Discovering ways to non-invasively early detect COVID-19 from heart rate, sleep, or a cough sound, even among the asymptomatic

November 6, 2020 | by

Heart rate, sleep quality, daily movement–cough sound frequency? Several studies in the US and UK are attempting to turn up ways to early diagnose mildly symptomatic, asymptomatic, or even pre-symptomatic COVID-19 cases, without the PCR swab or a blood test.

The more obvious of the two comes out of the Scripps Research Translational Institute. The DETECT study started in March (!) with 30,500 participants sending in data in the first six weeks of the study on heart rate, sleep quality, and daily movement. This information was then matched with self-reported symptoms and diagnostic tests taken if any. In this way, new infections and outbreaks could be detected at an earlier stage.  The study is attempting to confirm if changes in those metrics in an individual’s pattern can identify those even at a pre-symptomatic or asymptomatic stage. 3,811 reported symptoms, 54 reported testing positive, and 279 negative for COVID-19. The numbers seem small, but the analysis carries out that the combination of sensor and symptom data performed better in discriminating between positive and negative individuals than symptom reporting alone. The symptom data were taken from Fitbits and any device connected through Apple HealthKit or Google Fit data aggregators, then reported on the research app MyDataHelps. FierceBiotech, Nature Medicine (study)

Also using vital signs, back in August, Fitbit released early data on a 100,000+ study where changes in heart rate and breathing could detect about half of diagnosed cases at least one day to a week before diagnosis. Symptomatic cases were 1,100 in this sample. Heart rate and breathing were detected to become more frequent in the symptomatic, with the variability in time between each heartbeat dropping, resulting in a more steady pulse. The preferred tracking was at night during rest. However, there was a 30 percent false positive rate on the algorithm used, which is extremely high. FierceBiotech Related to this work, Fitbit was selected at the end of October by the US Army Medical Research and Development Command (USAMRDC) to receive nearly $2.5 million from the US Department of Defense through a Medical Technology Enterprise Consortium (MTEC) award to advance a wearable diagnostic capability for the early detection of a COVID-19 infection. Fitbit will be working with Northwell Health’s Feinstein Institutes for Medical Research to validate their early detection algorithm. Business Wire

And what about that ‘Covid Cough’? MIT is researching that this cough is different than other coughs, like from cold or allergy. Their research found that there’s a difference in the sound of an asymptomatic individual’s cough–and that sound frequency difference could not be heard by human ears. (Dog ears perhaps?) MIT researchers created “the largest audio COVID-19 cough balanced dataset reported to date with 5,320 subjects” out of 70,000 cough samples. The algorithm performed well. “When validated with subjects diagnosed using an official test, the model achieves COVID-19 sensitivity of 98.5% with a specificity of 94.2% (AUC: 0.97). For asymptomatic subjects it achieves sensitivity of 100% with a specificity of 83.2%.” This sure sounds like an AI screening tool that is inexpensive and convenient to use with multiple populations even daily. IEEE-EMB  BBC News reports that similar studies are taking place at Cambridge University, Carnegie Mellon University, and UK health start-up Novoic. The Cambridge study used a combination of breath and cough sounds and had an 80 percent success rate in identifying positive coronavirus cases from their base of 30,000 recordings.

All of these will be useful, but still need to be validated–and that takes time, for which this Editor thinks is short as this virus, like others, will eventually 1) mutate out or 2) be effectively treated as we do with normal flus. But down the road, these will serve as a template for new ways for early screening or even diagnosis of other respiratory diseases.

Perspectives: How Advanced Communications Technology Has Created A ‘New Normal’ In Healthcare

October 30, 2020 | by

TTA has an open invitation to industry leaders to contribute to our Perspectives non-promotional opinion area. Today, we have a contribution from Dave O’Shaughnessy, Avaya’s Healthcare Leader for EMEA and APAC, with a brief discussion of how AI and advanced communications technology can help healthcare in the long term. (It’s hard to say ‘a post-COVID world as France and Germany are experiencing second round lockdowns, and UK may not be far behind.) Interested contributors should contact Editor Donna. (We like pictures and graphs too)

Across industries, we see working patterns being transformed to create the ‘new normal’ as a result of COVID-19 and our reactions to the pandemic. The healthcare sector has been no different. The pandemic and its restrictions have brought a great number of new challenges to healthcare systems. And as has been the case across so many other sectors, communications technology has stepped in to plug the gaps caused by the pandemic.

The good news is that, not only have communications solutions successfully plugged the gaps, but they’ve also provided a blueprint for the future of healthcare. As we’ve found in other industries, we’ve actually seen the intelligent adoption of this technology lead to better experiences for patients, and better outcomes for providers, than were present before.

The most important (and immediate) area where this is most obvious is in contact tracing – tracking the physical, interpersonal interactions of those who have tested positive for COVID-19. This helps identify people who may need to be quarantined more quickly, therefore reducing the spread of the virus.

Helping government and healthcare organizations across the world with their contact tracing efforts, what we’ve found is that the most effective contact tracing efforts make use of artificial intelligence and automation. After all, the effort involves mountains of meticulous information gathering and analysis—all required to meet standards set by global health and government agencies. Acting upon that data manually just isn’t feasible, given the immediate needs at hand.

Therefore, the best systems employ AI virtual agents for initial patient contact, as well as for the simple data collection interactions – only falling back to live agents when the interaction becomes more complex. AI is also employed to deliver cloud-based, proactive notifications to automatically reach out to individuals or groups with optional response tracking, text interaction, and auto-forms to capture critical information.

Patients benefit from a smoother experience while providing the tracing information required, while healthcare providers and governments are able to collect more information with the resources they have.

Even without these focused AI technologies, however, our customers are putting their advanced contact centers to good use in combating the pandemic. In Saudi Arabia, for instance, one medical facility adopted a multi-experience approach, making it easy for patients to get the COVID-19-related information they need through a wide range of communications channels. This provided demonstrated results for improved knowledge on coronavirus safety measures in the community.

Going forward, we see tremendous use cases for extending this technology to make it easier for patients to directly engage with their doctors through asynchronous messaging. Such capabilities are of particular interest to mental health providers, who have found themselves unable to conduct in-person therapy sessions in the face of increased demand.

All of these solutions were implemented because of specific, pandemic-related challenges. But once the pandemic subsides, they’ll continue providing value, making it easier for patients to consume healthcare services, while delivering increased efficiency for providers.

Hat tip to Mary Burtt of AxiCom UK

Weekend ‘Must Read’: Are Big Tech/Big Pharma’s health tech promises nothing but a dangerous fraud?

July 24, 2020 | by

If it sounds too good to be true, it isn’t. And watch your wallet. In 14 words, this summarizes Leeza Osipenko’s theme for this article. It may seem to our Readers that Editor Donna is out there for clicks in the headline, but not really. Dr. Osipenko’s term is ‘snake oil’. It’s a quaint, vintage term for deceptive marketing of completely ineffective remedies, redolent of 19th Century hucksters and ‘The Music Man’. Its real meaning is fraud.

The promise is that Big Data, using Big Analytics, Big Machine Learning, and Big AI, will be a panacea for All That Ails Healthcare. It will save the entire system and the patient money, revolutionize medical decision making, save doctors time, increase accuracy, and in general save us from ourselves. Oh yes, and we do need saving, because our Big Tech and Big Health betters tell us so!

Major points in Dr. Osipenko’s Project Syndicate article, which is not long but provocative. Bonus content is available with a link to a London School of Economics panel discussion podcast (39 min.):

  • Source data is flawed. It’s subject to error, subjective clinical decision-making, lack of structure, standardization, and general GIGO.
  • However, Big Data is sold to health care systems and the general public like none of these potentially dangerous limitations even exist
  • Where are the long-range studies which can objectively compare and test the quality and outcomes of using this data? Nowhere to be found yet. It’s like we are in 1900 with no Pure Food Act, no FDA, or FTC to oversee.
  • It is sold into health systems as beneficial and completely harmless. Have we already forgotten the scandal of Ascension Health, the largest non-profit health system in the US, and Google Health simply proceeding off their BAA as if they had consent to identified data from practices and patients, and HIPAA didn’t exist? 10 million healthcare records were breached and HHS brought it to a screeching halt.
    • Our TTA article of 14 Nov 19 goes into why Google was so overeager to move this project forward, fast, and break a few things like rules.
  • We as individuals have no transparency into these systems. We don’t know what they know about us, or if it is correct. And if it isn’t, how can we correct it?
  • “Algorithmic diagnostic and decision models sometimes return results that doctors themselves do not understand”–great if you are being diagnosed.
  • Big Data demands a high level of math literacy.  Most decision makers are not data geeks. And those of us who work with numbers are often baffled by results and later find the calcs are el wrongo–this Editor speaks from personal experience on simple CMS data sets.
  • In order to be valuable, AI and machine learning demand access to potentially sensitive data. What’s the tradeoff? Where’s the consent?

Implicit in the article is cui bono?

  • Google and its social media rivals want data on us to monetize–in other words, sell stuff to us. Better health and outcomes are just a nice side benefit for them.
  • China. Our Readers may also recall from our April 2019 article that China is building the world’s largest medical database, free of those pesky Western democracy privacy restrictions, and using AI/machine learning to create a massive set of diagnostic tools. They aren’t going to stop at China, and in recent developments around intellectual property theft and programming back doors, will go to great lengths to secure Western data. Tencent and Fosun are playing by Chinese rules.

In conclusion:

At the end of the day, improving health care through big data and AI will likely take much more trial and error than techno-optimists realize. If conducted transparently and publicly, big-data projects can teach us how to create high-quality data sets prospectively, thereby increasing algorithmic solutions’ chances of success. By the same token, the algorithms themselves should be made available at least to regulators and the organizations subscribing to the service, if not to the public.

and

Having been massively overhyped, big-data health-care solutions are being rushed to market in without meaningful regulation, transparency, standardization, accountability, or robust validation practices. Patients deserve health systems and providers that will protect them, rather than using them as mere sources of data for profit-driven experiments.

Hat tip to Steve Hards.

Symptom checker K Health gains $48 million Series C (NY/Tel Aviv)

February 28, 2020 | by

While we’re on the subject of symptom checkers (Babylon Health below), K Health, a competitor in the US HQ’d in NYC, but also based in Tel Aviv, announced today their win of $48 million in a Series C funding round, led by 14W and Mangrove Capital Partners. Lerer Hippeau, Anthem (also a partner), Primary Ventures, and others participated. Their total funding is $97 million since November 2016. The new funding, according to Crunchbase News, will be used to scale the model, expand primary care to mobile devices, and expand to international markets. 

K (as they call themselves) concentrates on three areas. One is an AI-powered symptom checker that uses millions (they state) of anonymized medical records to provide a virtual consult. According to Crunchbase, the medical records came from Israel’s second-largest HMO, Maccabi, over 20 years. The app questions the user based on previous answers. K contrasts it to static protocols, or rules-based symptom checking. The second is to provide a primary care visit via text for $19/visit (or unlimited for $39/year) with free follow-ups over two weeks. The third is mental health, specifically treatment for anxiety and depression, a growing area both online and via mobile. The $29/month fee covers unlimited doctor visits and delivered prescription medication, excepting meds that require blood testing.

The symptom checker is available throughout the US and primary care in 47 states. According to Crunchbase’s interview with CEO Allon Bloch, they recently passed their 3 millionth user and are now available in Spanish. The company has grown in the past year from 80 to 200 people. Originally, the company linked to New York-based providers, but moved away from that to the primary care/text model. Their overall goal is to provide affordable diagnoses that are a lot more accurate than ‘Dr. Google’ and that steer the patient to the right care.

Should Babylon Health be serious about expansion to the US, they will be running up against K Health, as well as competitors such as 98point6. In the hybrid app-and-physical model, there are Carbon Health and One Medical. Also Mobihealthnews 

The confusion within TEC/telehealth between machine learning and AI-powered systems

September 25, 2019 | by

Defining AI and machine learning terminology isn’t academic, but can influence your business. In reading a straightforward interview about the CarePredict wearable sensor for behavioral modeling and monitoring in an AI-titled publication, this Editor realized that AI–artificial intelligence–as a descriptor is creeping into all sorts of predictive systems which are actually based on machine learning. As TTA has written about previously [TTA 21 Aug], there are many considerations around AI, including the quality of the data being fed into the system, the control over the systems, and the ability to judge the output. Using the AI term sounds so much more ‘techie’–but it’s not accurate.

Artificial intelligence is defined as the broader application of machines being able to carry out tasks in a ‘smart’ way. Machine learning is tactical. It’s an application that assumes that we give the machine access to data and let the machine ‘learn’ on its own. Neural networks in computer design have made this possible. “Essentially it works on a system of probability – based on data fed to it, it is able to make statements, decisions or predictions with a degree of certainty.”, as stated in this Forbes article by Bernard Marr.

CarePredict has been incorporating many aspects of machine learning, particularly in its interface with the wrist-worn wearable and its interaction with sensors in a residence. It gathers more over time than older systems like QuietCare (this Editor was marketing head) and with more data, CarePredict does more and progressed beyond the relatively simple algorithms that created baselines in QuietCare. They now claim effective fall detection, patterns of grooming and feeding, and environment. (Disclosure: this Editor did freelance writing for the company in 2017)

In wishing CEO Satish Movva much success, this Editor believes that using AI to describe his system should be used cautiously. It makes it sound more complicated than it is to a primarily non-techie, senior community administrative and clinical audience. Say what you do in plain language, and you won’t go wrong. AI for Healthcare: Interview with Satish Movva, Founder & CEO of CarePredict

 

Are AI’s unknown workings–fed by humans–creating intellectual debt we can’t pay off?

August 21, 2019 | by

Financial debt shifts control—from borrower to lender, and from future to past. Mounting intellectual debt may shift control, too. A world of knowledge without understanding becomes a world without discernible cause and effect, in which we grow dependent on our digital concierges to tell us what to do and when.

Debt theory and AI. This Editor never thought of learning exactly how something works as a kind of intellectual paydown of debt on what Donald Rumsfeld called ‘known unknowns’–we know it works, but not exactly how. It’s true of many drugs (aspirin), some medical treatments (deep brain stimulation for Parkinson’s–and the much-older electroconvulsive therapy for some psychiatric conditions), but rarely with engineering or the fuel pump on your car. 

Artificial intelligence (AI) and machine learning aren’t supposed to be that way. We’re supposed to be able to control the algorithms, make the rules, and understand how it works. Or so we’ve been told. Except, of course, that is not how machine learning and AI work. The crunching of massive data blocks brings about statistical correlation, which is of course a valid method of analysis. But as I learned in political science, statistics, sports, and high school physics, correlation is not causality, nor necessarily correct or predictive. What is missing are reasons why for the answers they provide–and both can be corrupted simply by feeding in bad data without judgment–or intent to defraud.

Bad or flawed data tend to accumulate and feed on itself, to the point where someone checking cannot distinguish where the logic fell off the rails, or to actually validate it. We also ascribe to AI–and to machine learning in its very name–actual learning and self-validation, which is not real. 

There are other dangers, as in image recognition (and this Editor would add, in LIDAR used in self-driving vehicles):

Intellectual debt accrued through machine learning features risks beyond the ones created through old-style trial and error. Because most machine-learning models cannot offer reasons for their ongoing judgments, there is no way to tell when they’ve misfired if one doesn’t already have an independent judgment about the answers they provide.

and

As machines make discovery faster, people may come to see theoreticians as extraneous, superfluous, and hopelessly behind the times. Knowledge about a particular area will be less treasured than expertise in the creation of machine-learning models that produce answers on that subject.

How we fix the balance sheet is not answered here, but certainly outlined well. The Hidden Costs of Automated Thinking (New Yorker)

And how that AI system actually gets those answers might give you pause. Yes, there are thousands of humans, with no special expertise or medical knowledge, being trained to feed the AI Beast all over the world. Data labeling, data annotation, or ‘Ghost Work’ from the book of the same name, is the parlance, includes medical, pornographic, commercial, and grisly crime images. Besides the mind-numbing repetitiveness, there are instances of PTSD related to the images and real concerns about the personal data being shared, stored, and used for medical diagnosis. A.I. Is Learning from Humans. Many Humans. (NY Times)

News roundup: docs dim on AI without purpose, ‘medtail’ a mall trend, CVS goes SDH, Kvedar to ATA, Biden ‘moonshot’ shorts out, and Short Takes

July 27, 2019 | by

Docs not crazy about AI. And Dog Bites Man. In Medscape‘s survey of 1,500 doctors in the US, Europe, and Latin America, they are skeptical (49 percent-US) and uncomfortable (35 percent-Europe, 30 percent-Latin America). Only 20 percent fess up to actually using an AI application, and aren’t crazy about voice tech even at home. Two-thirds are willing to take a look at AI-powered tech if it proves to be better than humans at diagnosis, but only 44 percent actually believe that will happen. FierceHealthcare

This dim view, in the estimation of a chief analytics and information officer in healthcare, Vikas Chowdhry, is not the fault of AI nor of the doctors. There’s a disconnect between the tech and the larger purpose. “Without a national urgency to focus on health instead of medical care, and without scalable patient person-centered reforms, no technology will make a meaningful impact, especially in a hybrid public goods area like health.” The analogy is to power of computing–that somehow when we focused behind a goal, we were able to have multiple moon missions with computing equivalent to a really old smartphone, but now we send out funny cat videos instead of being on Mars. (And this Editor growing up in NJ thought the space program was there to market Tang orange drink.) HIStalk.

Those vacant stores at malls? Fill ’em with healthcare clinics! And go out for Jamba Juice after! CNN finally caught up with the trend, apparent on suburbia’s Boulevards and Main Streets, that clinics can fill those mall spots which have been vacated by retail. No longer confined to ‘medical buildings’, outpatient care is popping up everywhere. In your Editor’s metro area, you see CityMDs next to Walmarts, Northwell Health next to a burger spot, a Kessler Health rehab clinic replacing a dance studio, and so on. The clever name for it is ‘medtail’, and landlords love them because they sign long leases and pay for premium spots, brighten up dim concourses, and perhaps stimulate food court and other shopping traffic. Of course, CVS and Aetna spotted this about years ago in their merger but are working expansion in the other direction with expanding CVS locations and on the healthcare side, testing the addition of social determinants of health (SDH) services via a pilot partnership, Destination: Health with non-profit Unite Us to connect better with community services. This is in addition to previous affordable housing investments and a five-year community health initiative. Forbes, Mobihealthnews

ATA announces Joseph Kvedar, MD, as President-Elect. Dr. Kvedar was previously president in 2004-5 and replaces John Glaser, PhD, Executive Senior Advisor, Cerner. He will remain as Vice President of Connected Health at Partners HealthCare and Professor of Dermatology at Harvard Medical School. A question mark for those of us in the industry is his extensive engagement with October’s Connected Health Conference in Boston, one of the earliest and now a HIMSS event. ATA’s next event is ATA2020 3-5 May 2020 in Phoenix–apparently no Fall Forum this year.

The Biden Cancer Initiative has shut down after two years in operation. This spinoff of the White House-sponsored ‘moonshot’ initiative was founded after the death of Beau Biden, son of Democrat presidential candidate Joe Biden. Both Mr. Biden and wife Jill Biden withdrew due to ethics concerns in April. According to Fortune, the nonprofit had trouble maintaining momentum without their presence. However, the setup invited conflict of interest concerns. The Initiative engaged and was funded by pharmas and other health tech companies, directly for Initiative support but mainly for indirect pledges to fund research. Most of these organizations do business with Federal, state and local governments. Shortly after the formal announcement, Mr. Biden the Candidate announced a rural health plan to expand a federal grant program to include rural telehealth for mental health and specialized services. Politico   But isn’t that already underway with the FCC’s Connected Care Pilot Program, coming to a vote soon? [TTA 20 June]

And…Short Takes

  • Philips Healthcare bought Boston-based patient engagement/management start-up Medumo. Terms not disclosed. CNBC
  • London’s Medopad launched with Royal Wolverhampton NHS Trust (RWT) in a three-year RPM deal. DigitalHealthNews
  • Parks Associates’ Connected Health Summit will be again in San Diego 27-29 August with an outstanding lineup of speakers. More information and registration here.

And in other news, Matt Hancock holds tight to his portfolio as UK Secretary of State for Health and Social Care in the newly formed Government under new PM Boris Johnson. Luckier than the other 50 percent!

 

 

Comings & goings: The TeleDentists go DTC, gains Reis as CEO; University of Warwick spinoff Augmented Insights debuts (UK); a new CEO leads GrandCare Systems

June 19, 2019 | by

The TeleDentists leap in with a new CEO. A year-old startup, The TeleDentists, has announced it will be going direct-to-consumer with teledentistry consults. This will permit anyone with a dental problem or emergency to consult with a dentist 24/7, schedule a local appointment in 24-48 hours. and even, if required, prescribe a non-narcotic prescription to a local pharmacy. Cost for the DTC service is not yet disclosed. Currently, the Kansas City-based company has provided their dental network services through several telehealth and telemedicine service providers such as Call A Doctor Plus as well as several brick-and-mortar clinic locations.

If dentistry sounds logical for telemedicine, consider that about 2 million people annually in the US use ERs for dental emergencies; 39 percent didn’t visit a dentist last year. Yet teledentistry is just getting started and is unusually underdeveloped, if you except the retail tooth aligners. Several US groups are piloting it to community health and underserved groups, with Philips reportedly considering a trial in Europe (mHealth Intelligence). This Editor notes that on their advisory board is a co-founder of Teladoc.  Release

The TeleDentists’ co-founder, Maria Kunstadter, DDS, last week announced the arrival of a new company CEO, Howard Reis. Mr. Reis started with health tech back in the 1990s with Nynex Science and Technology piloting telemedicine clinical trials at four Boston hospitals, which qualifies him among the most Grizzled Pioneers. He also was business development VP for Teleradiology Specialists and founding partner of The Castleton Group, a LTC telehealth company, and has worked in professional services for Accenture, Telmarc and SAIC/Bellcore. Most recently, he started teleradiology/telehealth firm HealthePractices. Over the past few years, Mr. Reis has also been prominent in the NY metro digital health scene. Congratulations and much success!  

In the UK, the University of Warwick has unveiled a spinoff, Augmented Insights Ltd. AI will be concentrating on machine learning and AI services that analyze long term health and care data, automating the extraction in real time of personalized, predictive and preventative insights from ongoing patient data. It will be headed by Dr. James Amor, whom this Editor met last summer in NYC. Long term plans center on marketing their analytics services to tech providers. Interested parties or potential users may contact Dr. Amor in Leamington Spa at James@augmentedinsights.co.uk |Congratulations to Dr. Amor and his team! 

And in more Grizzled Pioneer news, there’s a new CEO at GrandCare Systems who’s been engaged with the company since nearly their start in 1993 and in its present form in 2005. Laura Mitchell takes the helm as CEO after various positions there including Chief Marketing Officer and several years leading her own healthcare and marketing consulting firm. Nick Mitchell rejoins as chief technology officer and lead software developer. Founders Charlie Hillman remain as an advisor and Gaytha Traynor as COO. Their offices have also moved to the Kreilkamp Building, 215 N Main Street, Suite 130, in downtown West Bend Wisconsin. GrandCare remains a ‘family affair’ as this profile notes. Congratulations–again!

AI and machine learning ‘will transform clinical imaging practice over the next decade’

April 24, 2019 | by

The great challenges in radiology are accuracy of diagnosis and speed. Yet for radiology, machine learning and AI systems are still in early stages. Last August, a National Institutes of Health (NIH)-organized workshop with the Radiological Society of North America (RSNA), the American College of Radiology (ACR) and The Academy for Radiology and Biomedical Imaging Research (The Academy) kickstarted work towards AI. Their goal was to collaborate in machine learning/AI applications for diagnostic medical imaging, identify knowledge gaps, and to roadmap research needs for academic research laboratories, funding agencies, professional societies, and industry.

The report of this roadmap was published in the past few days in Radiology, the RSNA journal. Research priorities in the report included:

  • new image reconstruction methods that efficiently produce images suitable for human interpretation from source data
  • automated image labeling and annotation methods, including information extraction from the imaging report, electronic phenotyping, and prospective structured image reporting
  • new machine learning methods for clinical imaging data, such as tailored, pre-trained model architectures, and distributed machine learning methods
  • machine learning methods that can explain the advice they provide to human users (so-called explainable artificial intelligence)
  • validated methods for image de-identification and data sharing to facilitate wide availability of clinical imaging data sets.

Another aim is to reduce clinically important errors, estimated at 3 to 6 percent of image interpretations by radiologists. Diagnostic errors play a role in up to 10 percent of patient deaths, according to this report.

It is interesting that machine learning, more than AI, is mentioned in the RSNA materials, for instance in stating that “Machine learning algorithms will transform clinical imaging practice over the next decade. Yet, machine learning research is still in its early stages.” Radiology actually pioneered store-and-forward technology, to where radiology interpretation has been farmed out nationally and globally for many years. This countered a decline in US radiologists as a percentage of the physician workforce that started in the late 1990s and continues to today with some positive trends (Radiology 2015). Perhaps this distribution model postponed development of machine learning technologies. Also Healthcare Dive, RSNA press release  

Babylon Health’s expansion plans in Asia-Pacific, Africa spotlighted

April 18, 2019 | by

Mobihealthnews’ interview with Ali Parsa of Babylon Health illuminates what hasn’t been obvious about the company’s global plans, in our recent focus on their dealings with the NHS. For its basic smartphone app (video consults, appointments, medical records), Babylon last year announced a partnership with one of Asia’s largest health insurers, Prudential [TTA 18 Sept 18], licensing Babylon’s software for its own health apps across 12 countries in Asia for an estimated $100 million over several years. Babylon has also been active in Rwanda and now reaches, according to their information, nearly 30 percent of the population. There’s also a nod to developments with the NHS.

Parsing the highlights in Dr. Parsa’s rather wordy quest towards less ‘sick care’, more ‘prevention over cure’, and making healthcare affordable and accessible to everyone ’round the clock:

  • Asia-Pacific: Working with Tencent, Samsung and Prudential Asia through licensing software is a key component of their business. By adding more users, they refine and add more quality to their services. (Presumably they have more restrictions on the data they send to Tencent than what they obtain in China.)
  • Africa: How do you offer health apps in an economically poor country where only 5 percent of the population has a smartphone? Have an app that works for the 75 percent who have a feature phone. Babyl Rwanda has 2 million users–30 percent of Rwanda’s population–and completes 2,000 consultations a day. Babyl also works with over 450 health clinics and pharmacies. The service may also be expanded across East Africa, and may serve as a model for similar countries in other regions.
  • UK and NHSX: About the new NHS-formed joint organization for digital services, tech, and clinical care, Dr. Parsa believes it is ‘fantastic’ and that “it is trying to bring the benefits of modern technology to every patient and clinician, and aims to combine the best talent from government, the NHS and industry. Its aim, just like ours, is to create the most advanced health and care service in the world, to free up staff time and empower patients.” (Editor’s note:  NHSX will bring together the Department of Health and Social Care, NHS England and NHS Improvement, overseeing NHS Digital. More in Digital Health, Computer Weekly.)

China’s getting set to be the healthcare AI leader–on the backs of sick, rural citizens’ data privacy

April 17, 2019 | by

Picture this: a mobile rural health clinic arrives at a rural village in Jia County, in China’s Henan province. The clinic staff check the villagers, many of them elderly and infirm from their hard-working lives. The staff collect vital signs, take blood, urine, ECGs, and other tests. It’s all free, versus going to the hospital 30 miles away.

The catch: the data collected is uploaded to WeDoctor, a private healthcare company specializing in online medical diagnostics and related services that is part of Tencent, the Chinese technology conglomerate which is also devoted to AI. All that data is uploaded to WeDoctor’s AI-powered cloud. The good part: the agreement with the local government that permits this also provides medical services, health insurance, pharmaceuticals and healthcare education to the local people. In addition, it creates a “auxiliary treatment system for general practice” database that Jia County doctors can access for local patients. According to the WIRED article on this, it’s impressive at an IBM Watson level: 

Doctors simply have to input a patient’s symptoms and the system provides them with suggested diagnoses and treatments, calculated from a database of over 5,000 symptoms and 2,000 diseases. WeDoctor claims that the system has an accuracy rate of 90 per cent.

and 

Dr Zhang Qiaofen, in nearby Ren Zhuang village, says the system it has made her life easier. “Since WeDoctor came to my clinic, I feel more comfortable and have more confidence,” she says. “I’m thankful to the device for helping me make decisions.”

The bad part: The patients have no consent or control over the data, nor any privacy restrictions on its use by WeDoctor, Tencent, or the Chinese government. Regional government officials are next pictured in the article reviewing data on Jia County’s citizens: village, gender, age, ailment and whether or not a person has registered with a village health check. Yes, attending these health checks is mandatory for the villagers. 

What is happening is that China is building the world’s largest medical database, free of those pesky Western democracy privacy restrictions, and using AI/machine learning to create a massive set of diagnostic tools. The immediate application is to supplement their paucity of doctors and medical facilities (1.5 doctors per 1,000 people compared to almost double in the UK). All this is being built by an estimated 130 private companies as part of the “Made in China 2025” plan. Long term, the Chinese government gets to know even more intimate details about their 1.3 billion citizens. And these private companies can make money off the data. Such a deal! The difference between China’s attitude towards privacy and Western concerns on same could not be greater.  More on WeDoctor’s ambitions to be the Amazon of healthcare and yes, profit from this data, from Bloomberg. WeDoctor is valued at an incredible $5.5 billion. Hat tip to HISTalk’s Monday morning update.