Healthcare AI News 3/13/24

News

Sixteen large health systems and Microsoft form the Trustworthy and Responsible AI Network (TRAIN), which will share best practices, register the use of AI in clinical settings, provide AI outcomes measurements, and facilitate creation of a federated national AI outcomes registry. The group will work with OCHIN and TruBridge to make sure that all organizations, regardless of resources, can benefit from TRAIN’s work. The recently formed Coalition for Health AI (CHAI) is a partner and says that TRAIN will operationalize its principles in health systems.

Google launches Google Cloud Vertex Search for Healthcare, a set of search-and-answer tools that integrate with its MedLM, Healthcare Data Engine and Cloud Healthcare FHIR APIs. It offers medically-tuned search of EHR data for clinician presentation, configurable cloud APIs, the ability to answer clinician questions while understanding medical terminology, and providing links to source data points to provide transparency.

The government of China is working with private companies to develop a medically focused AI chatbot based on Meta’s Llama 2.0. It is testing an AI assistant model for neurosurgeons at seven hospitals in Beijing. The CARES Copilot 1.0 was trained on medical literature and can process images, audio, text, MRI, ultrasounds, and CT scans.

Microsoft previews healthcare-specific safeguards for Azure AI Health Bot services that integrate with Copilot Studio to allow healthcare organizations to develop their own copilots.

Business

Symplr announces SymplrAI, which will offer healthcare efficiency tools via Amazon Web Services AI/ML.

Peter Thiel-backed Cognition AI unveils Devin, an AI-powered software engineer that can generate, debug, and deploy code for websites and videos from a single command. It has its own command line, code editor, and browser.

Amazon Web Services lists the generative AI solutions that its healthcare customers have created:

• Digital pathology (Philips).
• Phone-based, voice-enabled progress note generation (Pieces Technologies).
• Hospital operations efficiency (Symplr).
• Enterprise imaging (Konica Minolta).

Biotech startup Insilico Medicine describes how it used its AI platform to create a drug candidate for idiopathic pulmonary fibrosis that is undergoing Phase II clinical trials. The company says the process took 2.5 years versus the six years that would have been needed without AI. The company used biology AI to discover the drug, then generated the molecule using chemistry AI.

Research

In England, University of Oxford researchers warn clinicians that using ChatGPT and Bard to create care plans could compromise patient privacy or mislead users with incorrect information.

In Israel, Sheba Medical Center researchers develop an AI platform to help mental health professionals diagnose and manage conditions efficiently, concluding that its diagnostic accuracy is comparable to that of experience psychiatrists.

A study finds that large language models can translate discharge summaries into readable, patient-friendly language, which helps providers comply with requirements to give patients immediate access to their discharge notes. However, the authors found examples of key information being omitted, most often history of present illness and procedures, which they attribute to prompt engineering that optimized readability in the form of shorter documents.

Other

CommonSpirit Health EVP/CIO Daniel Barchi, MS says that healthcare’s adoption of AI should follow the Maslow’s Hierarchy of Needs in first being applied to technical needs, then moved up the triangle to improve healthcare efficiency and safety. (click graphic to enlarge)

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 3/6/24

News

UPMC physicians develop a AI-powered smartphone app that can diagnosis acute otitis media by analyzing video from a phone-connected otoscope.

Pediatric clinicians at Mass General Brigham create a series of 45 instructional smartphone videos for clinicians, then use ChatGPT to create Spanish language versions that they distributed to nurses in Guatemala and Colombia.

The Coalition for Health AI names co-founder Brian Anderson, MD as CEO and co-founder John Halamka, MD, MS as board chair. The company added federal leaders to its board and will work with the government to develop quality and safety standards.

The American Medical Association’s house of delegates defines AI as “augmented intelligence” rather than “artificial intelligence” to focus on AI’s role of enhancing human intelligence rather than replacing it.

Business

UMass Memorial Health will use Google Cloud to build patient care and research tools.

Health and human services software vendor Healthy Together acquires Kinsa Health, which offers AI-powered epidemiological prediction models using consumer smart thermometers as well as healthcare demand forecasting. Kinsa had shut down several months ago after 12 years, with founder and CEO Inder Singh expressing hope that he could find a new home for the company’s work.

Machinify launches an AI-powered healthcare claims system that includes apps for prior authorization approval, claims auditing, and claims review and correction. Founder and CO Prasanna Ganesan, PhD was a co-founder of digital video store Vudu, which was acquired by Walmart in 2010 and then in 2020 by Fandango, which now brands it as Fandango At Home.

Research

The federal government awards Philadelphia-based non-profit Every Cure a $48 million grant to use AI to find new uses for existing drugs. Penn medical school professor and immunologist David Fajgenbaum, MD, MBA, MSc co-founded the organization after saving his own life by finding a “repurposed” drug, a process he described in his book.

Other

Attorneys list potential areas of liability exposure for using healthcare-related AI, urging providers to avoid Practice Fusion type kickback situations by vetting the product’s explainability and the governance and privacy practices of its vendor:

• Payers using AI to manage prior authorization while denying legitimate claims or improperly influencing the process.
• Using AI to analyze medical images that could raise questions about how the system were trained, whether their vendors are paid for the volume and value of the resulting referrals, and kickback implications of AI systems that are tied to specific treatments.
• Possible clinical studies fraud in which AI could be tweaked to overstate drug efficacy.

Open Notes launches OpenNotes Lab, which will advocate and study the use of AI to enhance trust and communication between patients and their care teams. The company is interesting in partnering with vendors, health systems, patient advocacy groups, professional associations, and regulatory agencies.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 2/28/24

News

Hartford HealthCare creates the Center for AI Innovation in Healthcare and launches Holistic Hospital Optimization, a startup company that will help hospitals optimize patient flow and operations.

A survey that was commissioned by Amazon Web Services finds that US healthcare employees and employers expect AI to deliver a 44% improvement in labor productivity, with potential uses in fraud detection, quality assurance, medical imaging review, and improving communication.

Highmark Health will use Epic’s Payer Platform running on Google Cloud to glean insights that can be used to inform consumers of the next best actions. The company says that more than half of its 7 million members are attributed to an Epic-using provider, allowing it to close care gaps automatically.

The VA says that 40 of the 100 AI use cases that it has identified are being used in the field.

Google CEO Sundar Pichai says that the company has taken its Gemini image generation feature offline for further testing after reports that it created offensive and biased results. Critics noted that the tool seemed exhibit liberal bias, such as creating an image of “a founding father of America” as black and Native American men.

Business

AI-powered clinical documentation tools vendor Abridge raises $150 million in a Series C funding round, valuing the six-year-old company at $850 million.

Research

Vanderbilt University Medical Center uses ChatGPT to analyze the medical literature to find existing drugs that could be repurposed for treating Alzheimer’s disease. The results were then applied to VUMC’s patient data and the All of Us Research Program to identify three drugs that may hold promise – losartan, metformin, and simvastatin.

Other

Scientists warn that while it’s possible to recreate a deceased loved one based on their emails and other writings, that could harm mental health, as it would impede the natural grieving process and create dependence on the technology to the exclusion of forming new relationships.

Another lawyer gets in trouble for using ChatGPT, as a law firm partner bills the court for $113,500 at the ChatGPT-recommended rate of $550 per hour. The judge halved the bill, saying that using ChatGPT to calculate the firm’s fee without submitting supporting detail is “utterly and unusually unpersuasive.”  

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 2/21/24

News

Tufts Medicine (MA) launches its Virtual-First AI initiative with a new digital assistant available through its MyTuftsMed app. The health system will add on-demand chat capabilities later this year.

Authenticx develops a generative AI tool that summarizes healthcare consumer conversations from multiple sources to offer insights into emerging issues and overall customer experiences.

Business

Fabric raises $60 million in a Series A funding round. The company offers AI-powered healthcare enablement software for virtual and in-person visits. It acquired AI-based virtual assistant business Gyant earlier this month.

UnityAI, an AI-powered hospital bed management software startup, raises $4 million in seed funding.

Automated medical coding and documentation software startup RapidClaims launches with $3.1 million in funding. Beta users say they’ve reduced claim denials related to coding errors by 70%.

ChatGPT developer OpenAI reportedly completes a deal that values it at $80 billion.

ChatGPT users, meanwhile, were left scratching their heads for several hours Tuesday night as the AI tool generated nonsensical responses well into the night.

Research

New research out of UPMC’s Center for Connected Medicine finds that very few health systems have formal policies in place regarding AI, and even fewer have frameworks in place for the use of generative AI. Seventy percent say they plan on adopting AI solutions offered by their EHR vendors, with improving efficiencies, automating repetitive tasks, and bringing more visibility to clinical decision-making noted as top reasons for implementation.

Other

Hartford Healthcare (CT) debuts its Center for AI Innovation. Through collaboration with teams from MIT and the University of Oxford in England, the center will focus on research, innovation, education, and invention. Current projects include using AI to predict length of stay and patient deterioration, as well as to optimize nurse scheduling and surgical efficiency.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 2/14/24

News

DeepScribe announces GA of its new Trust and Safety Suite, a set of solutions designed to give users greater visibility into the safety and reliability of its AI-powered medical scribe software.

Persistent Systems develops AI-powered population health management software that identifies social determinants of health based on EHR data and then recommends personalized care interventions.

CitiusTech develops the Gen AI Quality & Trust framework and support service to help healthcare software developers design, implement, and scale vetted AI solutions across enterprise environments.

Business

Anatomy Financial launches with $7.6 million in funding. The San Francisco-based company has developed AI-powered financial automation software for medical and dental practices, and digital health and healthcare billing companies.

Children’s Hospital of Los Angeles implements Vital’s AI-powered ERAdvisor software as a part of its MyVisit app.

Research

A series of studies presented at the annual meeting of American Academy of Orthopaedic Surgeons determine that the accuracy of musculoskeletal health information provided by chatbots used by ChatGPT, Google Bard, and BingAI is limited. The chatbots displayed especially significant limitations when asked to provide clinical management suggestions, omitting important steps such as ordering antibiotics before cultures.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 2/7/24

News

Amazon announces Rufus, an AI tool that is trained on the company’s product catalog and on web information to answer customer questions about products, comparisons, and recommendations.

Google will reportedly rename its Bard AI tool to Gemini this week.

Attorneys say that the legal liability frameworks of the US and the EU are not equipped to determine who will be responsible for medical injuries that are caused by black box AI systems.

Business

UC Davis Health chooses Viz.ai to help diagnose strokes from CT scans for quick intervention.

Radiology AI marketplace CARPL raises $6 million in seed funding.

Laguna Health, which uses AI to create personalized care plans that incorporate social, emotional, and cultural factors, raises $15 million in a Series A funding round.

Ambience Health raises $70 million in a Series B funding round that was led by Kleiner Perkins and OpenAI Startup Fund.

Research

Researchers find that their AI model that predicts how well schizophrenia patients will respond to antipsychotic medications performed well, but didn’t work when used in other studies. Experts warn that AI’s ability to advance medicine is limited by the methods that are used to collect, process, store, and analyze data. A co-author of the study predicts that it will be more than 10 years until AI will be able to analyze patient questionnaire results well enough for clinical use in psychiatry.

The VA is using smart rings and AI to collect the health data of veterans who were exposed to open-air burn pits during wars in Iraq and Afghanistan in a study that is being conducted through Northwell Health. Researchers hope to correlate vague symptoms with monitored data and screening questionnaires.

Other

AHIMA launches AI Resource Hub, which will provide health information professionals with information about non-clinical use of AI in healthcare.

A New York lawyer faces disciplinary action for citing non-existent cases in medical malpractice lawsuit documents, as provided by and hallucinated by ChatGPT without being checked by the lawyer.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 1/31/24

News

Oracle announces a generative AI service that supports text generation, text summarization, and LLM-contextualized search results. The company says it will release search, aggregation tools, and prebuilt agent actions across its SaaS applications, including those of Oracle Health.

Apple will reportedly release AI features with IOS 18 in June, reportedly including enhancing Messages to answer questions and auto-complete sentences.

The State of Illinois launches a Google AI-powered portal for families to access behavioral and mental health resources.

In UAE, Emirates Health Services launches an AI-powered virtual nurse that answers patient questions, provides medical advice, and helps with assessment and triage. EHS will also assess patient-reported symptoms to schedule appointments with the correct specialty. EHS is running a long list of AI projects, including those related to diagnosis, treatment, disease prediction, patient monitoring, and speech-powered clinical documentation.

Microsoft and Google warn investors that the arms race to develop AI products will be expensive due to the computing power it requires.

Business

The Justice Department is reportedly issuing subpoenas to drug companies and EHR vendors to determine if AI is being used to influence prescribing in ways that breach anti-kickback and false claims violations

Microsoft and OpenAI are reportedly in talks to invest up to $500 million in AI-powered robot developer Figure AI, which could value the 2022 startup at $2 billion. The latest autonomous model is 5’6” tall, can carry 44 pounds, walks at 1.2 meters per second, and operates for five hours when charged.

Research

Researchers apply AI and sophisticated analytics methods to test the detection of Medicare fraud, noting that traditional auditors only have enough time to look for specific suspicious patterns.

Other

Cleveland Clinic says in its annual meeting that it is piloting AI for answering the questions of patients with chronic disease, creating ambient clinical documentation, predicting patient volume, and creating draft responses to questions that are submitted via Epic’s MyChart.

The hottest job in corporate America is the executive in charge of AI, according to the New York Times. It notes that Mayo Clinic in Arizona has hired its first chief artificial intelligence officer. 

An AAMC story describes the experience of physicians who are using AI to create documentation from recordings of their patient encounters. Eric Poon, MD, MPH (Duke Primary Care) says he is finishing his clinical schedule on time for the first time in his life, also noting that he didn’t realize how his EHR typing was slowing down the visit. Matthew Anderson, MD (Atrium Health) says he has a summarydraft in his hand 15 seconds after the visit ends. However, users say that ambient listening tools aren’t perfect in deciding what parts of the conversation to include or exclude, and some of them report that they spend quite a bit of time editing out extraneous information. They also say that it works well for quick visits, but less so when visits involve multiple medical issues.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 1/24/24

News

WHO publishes recommendations for governments on using AI for public health and medical purposes:

• Give developers access to computing power and paid data sets if they agree to follow ethical principles.
• Use laws and regulations to ensure that health-related use meets ethical obligations and human rights standards.
• Assign a regulatory agency to assess and approve health AI applications.
• Require third-party post-release auditing and impact assessments.
• WHO also recommends that AI developers include potential users and stakeholders in their work, including patients, and design their products to perform well-defined tasks that can improve health system capacity and patient interests.

Amazon is reportedly developing Alexa Plus, a GPT-enhanced version of its voice assistant that will be able to conduct conversations similar to ChatGPT, but with a subscription required to help the company cover the costs of running the AI models. Insiders who are involved with a limited preview say that the new Alexa is good at conducting conversations, but it still gives overly long or incorrect responses and sometimes fails to provide the desired answer. Insiders worry whether consumers will prove willing to pay for a new Alexa version given the market struggles of the current one.

OpenAI signs its first partnership with a university, in which Arizona State University students and faculty will use ChatGPT Enterprise to create personalized AI tutors and to expand their work with prompt engineering.

Google adds experimental AI features to its Chrome browser for tab organization, theme creation, and a “help me write” feature for posting online comments or reviews.

In India, Apollo Cancer Centre in Bangalore launches the country’s first AI-Precision Oncology Centre, which will use AI for diagnosis, risk assessment, development of treatment protocols, identifying patients for targeted therapy and immunotherapy, and patient and family education.

Business

Amazon lists the ways that Amazon Pharmacy is using AI:

• Converting unstructured prescription data into standard categories, such as dose and frequency, to speed up prescription processing by 90%, reduce errors, and to provide clear patient instructions.
• Forecast drug demand to make decisions about which medications to stock and where to store them.
• Provide customers with real-time insurance estimates and look for better pricing within Amazon’s programs such as Prime savings, RxPass, or automatic coupons.
• Help team members answer questions by summarizing internal documentation and knowledge bases.
• Batch refills for efficient filling based on when the prescription is needed.

Research

Cedars-Sinai investigators develop an AI model to predict outcomes for pancreatic cancer patients using biomarkers. Developing the model, called Molecular Twin, also provided new insight into the high predictive value of plasma proteins. They say that their low-resource “parsimonious model” approach, which uses commonly available pathology specimens and clinical data, can impact clinical care and democratize the use of precision medicine for treating cancer.

Other

The business school of Washington University in St. Louis will offer a symposium titled “AI and Innovation in Healthcare” on Thursday. A keynote speaker is Centene CEO Sarah London, MBA, whose background includes healthcare analytics and revenue cycle jobs with R1 RCM, Humedica, and Optum. The event is free for either virtual or in-person attendees.

In India, scammers use AI to impersonate the voice of a woman’s brother-in-law, convincing her to send them a mobile phone payment $400 to pay his daughter’s emergency hospital bill.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 1/17/24

News

Nvidia founder and CEO Jensen Huang tells attendees of the J.P. Morgan Healthcare Conference that he expects drug design to move nearly completely to AI systems. He also predicts that medical devices such as ultrasound and CT scanners will be “a device plus a whole bunch of AIs” and that will create incredible value and opportunities.

A UC Berkeley public health professor and psychiatrist warns that using AI to offset a shortage of mental health providers and to address growing problems such as loneliness raises concerns. Jodi Halpern, MD, PhD  worries about the effect of companies marketing AI conversational bots as therapists and trusted companions to people who are depressed or vulnerable. She says that unlike therapists, app vendors are not regulated, companies may be motivated to use the “addictive engineering” of social media companies, and app users may rely on the app the point of withdrawing from human engagement.

A Georgia lawmaker introduces a bill that would make it illegal to use AI to discriminate against people or for making health insurance and public services decisions without human review.

Family-powered autism therapy AI platform provider Forta raises $55 million in a Series A funding round.

A Google-created AI system called AMIE that was optimized for diagnostic dialog outperforms primary care doctors in synchronous text chat with patient actors. The authors warn that while synchronous text chatting is not a common practice, their work is a milestone in developing conversational diagnostic AI to take patient histories, provide diagnostic support, and communicate with skill and empathy.

Business

Investor Mark Cuban predicts that AI will transform healthcare, but will involve millions of models that are created by individual companies and providers instead of a handful of big-name models. He says that the biggest issue is intellectual property protection, and expects health systems to limit access to the models they develop.

Workers’ compensation insurer MEMIC will use AI technology from Clara Analytics to automate medical records transcription and data extraction for claims reviews.

The CEO of medical device maker Medtronic says that the company will transform its products with AI, giving recent examples of its solutions for intelligent endoscopy, digital surgery, and remote surgery assistance.

Research

Mass General Brigham researchers show that large language models can be trained to automatically extract social determinants of health from clinician notes. The finely tuned model identified 94% of patients who have adverse SDoH versus the 2% that could be identified using diagnostic codes. The model was trained to identify sentences that refer to employment status, housing, transportation, relationships, the availability of social support, and parental status. Bias was less likely than with GPT-4. The model was much smaller than ChatGPT models and results were better even thought the model was trained on just a few hundred patient records.

Switzerland-based cardiologists and researchers use AI to improve the usefulness of the atrial fibrillation detection capabilities of consumer wearables that are made by AliveCor, Apple, Fitbit, and Samsung. The study authors say that the single-lead ECG functionality alone is not clinically useful because of high numbers of inconclusive tracings, but the model reduced the percentage of those from 16% to 1.2%. They caution that opportunities for further studies will be limited as those manufacturers enhance their algorithms in undisclosed ways and may restrict access to their raw data. The study used cardiac diagnostics AI products from PulseAI.

A Michigan Medicine team uses AI to review physician notes for signs of risky drinking, which identified three times the number of patients than would have been flagged by diagnosis codes alone.

ChatGPT failed to make an accurate diagnosis in journal-published pediatric case challenges 83% of the time. The authors note that systems such as Google’s Med-PaLM 2 that were trained on medical data would likely perform better.

Pharmacy professors find that ChatGPT performs well in answering questions that patients might ask a pharmacist who is working remotely. It also did a good job of answering general patient questions, offering dietary suggestions, encouraging patient adherence, and clarifying medical terms.

Other

UCSF Health is developing AI tools for nurses, which include patient deterioration detection, patient placement, and matching patients with specialized nurse expertise. A representative of UCSF Health’s nursing union isn’t convinced, declaring that AI has created a “disaster-capitalism moment” in being used to improve efficiency with an end result of reducing access to skilled care.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 1/10/24

News

OpenAI launches the GPT Store, which allows users to distribute – and eventually charge for – customized GPTs that they have created. The company also announces ChatGPT Team, a $25 per-user per-month secure workspace for teams of up to 149 people that supports GPT sharing. The company lists Boston Children’s Hospital as a early adopter, which says that it has used Team to pilot GPTs for productivity and collaboration.

OpenAI announces that wearables vendor Whoop has incorporated GPT-4 into its app that offers personalized fitness and health coaching related to heart rate, workouts, sleep, and stress. The Whoop 4.0 device is free for those who sign up for a $239 per year subscription.

Microsoft changes the layout of its keyboard for the first time in 30 years by adding a Copilot key that invokes the AI assistant.

Business

The Mayo Clinic will use computing chips and systems from Cerebras to develop its own AI models, such as summarizing a patient’s medical records, analyzing diagnostic images, and reviewing genomic data.

Sword, which offers digital health solutions and services for pain prevention and treatment, says that it delivered 1.5 million AI Care sessions in 2023. Sword says that it developed and patented the world’s first AI care solution in 2018.

ShifMed launches ShiftAdvisor, an AI-powered solution that optimizes nurse shift scheduling by considering preferred days, times, pay, and locations.

Accenture invests in Israel-based QuantHealth, which uses AI to simulate drug clinical trials in the cloud.

Innovaccer announces Sara, an AI scribing assistant that transcribes, summarizes, and analyzes visit conversations to create SOAP notes.

Direct-to-consumer drug marketing technology vendor Swoop enhances its AI drug adherence tool to allow drug companies to market their products to patients and their doctors before a diagnosis has been made. They don’t say how they do that, but its other products create custom market segments using the de-identified data of 300 million people that it gets from Datavant and Epsilon.

Other

Vanderbilt University Medical Center lists its AI accomplishments in clinical and research settings:

• Running a 10-physician pilot of Nuance DAX Copilot to create encounter notes from ambient listening.
• Offering researchers a VUMC-created, GUI-based version of OpenAI’s large language module, which is HIPAA certified under VUMC’s business associate agreement with Microsoft. VUMC has disabled access to the public version of ChatGPT on its networks.
• Creating the IQueue platform to optimize chemotherapy infusion appointments.
• Developing a stroke patient evaluation tool that delivers quantified and color-coded CT perfusion maps.
• Creating its own AI application to predict elective surgical case volume.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 1/3/24

News

Microsoft rolls out the Copilot app, which was formerly known as Bing Chat, for Android and IOS. It provides subscription-free access to OpenAI’s GPT-4 and the DALL-E3 image generator.

Humetrix announces a cloud-based international health communicator platform that presents clinicians with a patient’s medical records that it has translated into the clinician’s own language.

VideaHealth receives FDA 510(k) clearance for an AI-powered diagnostic tool for dental diseases, which will also be distributed by Henry Schein One to its dental software users.

In the Netherlands, the CMIO of University Medical Centre Groningen urges the EU to avoid overregulating AI since “it might be the only chance to have some level of healthcare shortly for older people.” The hospital is using AI tools to create draft responses to patient emails and to summarize the information in Epic for rounding,

Business

Bain & Company believes that investment in healthcare-related generative AI is just starting and expects the technology to drive productivity gains, offer a better provider and patient experience, and improve outcomes. It says that companies that were built around AI are raising significant investments, but also that the private equity owners of mature companies are adding AI to improve their products or businesses, which may pose a disruptive threat to some portfolio companies. The highest-risk business are those that provide services for content generation, administrative processes, call centers, and text writing and summarization.

Health insurance brokers are using AI to simplify the process of shopping for ACA marketplace and Medicare Advantage plans. The tools collect an individual’s information, make predictions about their health needs, and then score available plans based on income, prescriptions, and preferred doctors. They can also benefit users by eliminating the incentive for brokers to push the highest-commission plan.

WellSpan Health touts the use of AI in its six hospitals, which includes Artisight virtual sitting and nursing technology, Aidoc for image review, and DAX for ambient documentation.

Research

Dartmouth medical educators develop AI Patient Actor, a ChatGPT virtual patient that medical students can interview while reviewing their vital signs and lab results. The app provides individualized feedback that the student can use to try again. The researchers expect the tool to provide a stress-free environment for students to practice their clinical interaction before moving on to clinical settings or working with actors who pose as standardized patients.

Other

A Toronto newspaper profiles ChartWatch, an AI-driven early warning system for patient deterioration that was developed by St. Michael’s Hospital. The hospital says that use of the risk scores, which are calculated hourly, was associated with a 26% mortality reduction among non-palliative patients in its general medicine unit. The hospital is spending $4 million annually to test AI solutions, about 50 so far. It has also developed solutions to help assign nurses to ED roles, display wait times in the ED, and summarize the medical records of patients with multiple sclerosis that span years. The reports note that Health Canada has approved only AI-powered software in which the algorithms are locked, which requires applying for a license amendment if the algorithms can learn or can be changed.

UCSD Health’s newly named chief AI officer Karandeep Singh, MD, MMSc says that the best near-term use of AI in healthcare involves “keyboard liberation” in drafting patient message replies, summarizing chart documentation, and creating document from ambient conversations. He predicts that AI-powered decision-making “is very far away.”

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 12/27/23

News

Researchers predict that AI will enhance the longer-term benefits of using wearables data, such as step tracking, either by directly providing health coaching or by analyzing wearables data to give human health coaches a better picture of the user’s health. The Time article observes that Google will add AI insights to users of its Fitbit devices, Google DeepMind is working on “life adviser,” and Apple will reportedly release an AI health coach next year. Experts suggest that while these projects are interesting, evidence that AI-analyzed wearables data has not been proven to improve outcomes or mortality.

AI researchers obtain the email addresses of 30 New York Times employees by feeding ChatGPT some known addresses and then asking for more via its API, which bypasses some of ChatGPT’s privacy restrictions. The article notes that AI companies can’t guarantee that their systems haven’t learned sensitive information, although AI tools are not supposed to recall their training information verbatim. Training on inappropriately disclosed medical records weren’t mentioned in the article, but should be concerning.

Business

OpenAI is reportedly discussing a new funding round that would value the company at or above $100 billion, making it the second-most valuable US startup behind SpaceX.

Research

Researchers say that while AI can help nurses by automating routine tasks and providing decision support, it cannot replace their excellence in critical thinking, adapting to dynamic situations, advocating for patients, and collaborating. The authors note that nurses have hands-on clinical experience in assessing and managing patient conditions; take a holistic approach that considers the physical, emotional, and psychological aspects of patient care; and requires them to make ethical and moral decisions that respect the patient’s values, beliefs, and culture.

A study finds that ChatGPT 3.5 does a good job in simplifying radiology reports for both clinicians and patients while preserving important diagnostic information, but is not suitable for translating those reports into the Hindi language.

Three Stanford graduate students develop a tool that can accurately guess where a photo was taken by using AI that was trained on Google Street View. The system was trained on just 500,000 images of Google Street View’s 220 billion but can guess the country 95% of the time and can usually guess the location of a photo within 25 miles. The system beat a previously undefeated human “geoguessing champion.” Civil liberties advocates worry that such a system, especially if it is ever rolled out widely by Google or other big tech firms, could be used for government surveillance, corporate tracking, or stalking.

Other

An expert panel that was convened by AHRQ and the National Institute on Minority Health and Health Disparities offers guiding principles for preventing AI bias in healthcare:

• Promote health and healthcare equity through the algorithm’s life cycle, beginning with identifying the problem to be solved.
• Ensure that algorithms and their use are transparent and explainable.
• Engage patient and communities throughout the life cycle.
• Explicitly identify fairness issues and tradeoffs.
• Ensure accountability for equity and fairness in AI outcomes.

Clinical geneticist and medical informaticist Nephi Walton, MD, MS warns that AI is convincing even when it is wrong. He asked ChatGPT how to avoid passing a genetic condition to his children and it recommended that he avoid having children. He says AI has improved, but the way that it is trained is a problem because it pushes old evidence and guidelines to the top while neglecting new information.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 12/20/23

News

OpenAI posts a guide to GPT prompt engineering, explaining six strategies with examples:

• Write clear instructions. 
• Provide reference text.
• Split complex tasks into subtasks.
• Give the model time to think.
• Use external tools.
• Test changes systematically.

Dictionary.com names the AI-related usage of “hallucinate” as its word of the year for 2023, the same year it was added to the dictionary.

Business

Healthcare ambient AI company Nabla announces that 10% of its clinician users were taking advantage of its recently released Spanish translation capability within its first four months, with especially high adoption among psychiatrists, psychologists, and psychotherapists. Two of the company’s three co-founders are former Facebook AI research engineers.

Healthcare payment integrity company Trend Health Partners acquires Advent Health Partners, which uses AI to assemble and present medical records for denials management, utilization management, bill reviews, and DRG reviews.

High-profile investor Vinod Khosla tells conference attendees that AI will impact healthcare by nudging patients to perform preventive health activities. He also says that clinicians practice “crude medicine” by using the same medications on every patient despite their unique genetic characteristics, predicting that medicines will be tailored to individual patients.

Healthcare AI vendor John Snow Labs deploys a medical chatbot using its own large language model running Oracle Cloud Infrastructure. The model provides detailed answers with citations to its research sources, serving as a personal research assistant to the latest biomedical literature or personal document repositories.

Research

A small study finds that AI can accurately screen people for autism spectrum disorder, and possibly determine its level of severity, from photos of their eyes.

Other

Eric Topol, MD of Scripps Research Translational Institute says that with cancer screening, “we’ve got this all wrong” because only 12 to 14% of cancers are being diagnosed via mass screening at a cost of tens or hundreds of billions of dollars each year. He adds that false positives create patient anxiety and it’s “so dumb” that the screening is based on age and misses major cancers among younger people. He says that AI can define individual risk by analyzing clinical notes, test results, and polygenic risk scores.

HHS’s use of AI ranks it #4 among federal government agencies, with projects that include:

• Generating clinical study reports using Phase I and II study data.
• Providing first responders with an app that tell them how many Medicare beneficiaries in a given area use electricity-dependent medical devices.
• NIH’s use of an AI tool to rank incoming grant applications to highlight those that address high-priority topics.
• CDC’s use of AI for surveillance testing.
• According to a GAO report, HHS is using an AI chatbot to provide automated email responses for physical security questions.

Bill Gates recaps 2023 by saying that it was the first time he used AI for serious work and “not just to mess around and create parody song lyrics for my friends,” predicting that AI will be used by a significant part of the US population within 24 months. He expects that in healthcare, AI will accelerate discovery of new drugs, predict antibiotic resistance and recommend optimal therapy, serve as a co-pilot for managing high-risk pregnancies, assess risk for HIV, and creating a voice-powered standardized patient record in Pakistan.

Mass General Brigham asks its experts for their 2024 predictions, many of which involved AI:

• Personalizing treatment plans, surgical precision, and post-operative monitoring in neurosurgery.
• Integrating AI into radiology practices.
• Using AI Improvements that allow efficient updating for safety and effectiveness.
• Using AI chatbots for initial patient triage.
• Microsoft, Amazon, and Alphabet acquiring key health AI players as the overall market consolidates.
• Deploying AI-inspired robots to support home caregivers for people with spinal cord injury.
• Using digital twins.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 12/13/23

News

Elon Musk’s AI firm XAI releases its Grok chatbot to Premium+ subscribers of his X social media platform. Premium+ costs $16 per month and in addition to Grok access, includes editing of X posts, the ability to write longer posts, reply prioritization, the blue user checkmark, and removal of ads.

The bioinformatics department of Yale’s medical school launches a 16-week, $5,000 online non-credit certificate program for medical professionals titled “Medical Software and Medical Artificial Intelligence.”

Emory University’s two-day symposium on healthcare AI draws 450 attendees.

Google releases MedLM, two medically tuned Med-PaLM models that are being piloted by HCA Healthcare for drafting physician notes and by BenchSci for biomarker classification. MedLM is available to US users of Google Cloud who are whitelisted through its Vertex AI.

Business

The Federal Trade Commission and UK regulators will review any anti-trust issues that may have resulted from Microsoft’s investment in OpenAI.

Drug maker Sanofi invests $140 million in AI-powered molecule development startup Aqemia.

Darena Solutions, Leidos, and SLI Compliance launch a certification program for AI applications that are built using SMART on FHIR.

Children’s Health (TX) rolls out a generative AI solution from Pieces Technologies that creates a 100-word summary of a patient’s visit using information extracted from Epic.

Other

Half of surveyed nursing students use generative AI tools to complete their assignments or tests, but just one-third of them think that it’s OK for colleges to use AI to make admission decisions.

The owner of Sports Illustrated fires the publication’s CEO, COO, media president, and corporate counsel for unstated reasons, two weeks after it was caught using AI-generated author bios and photos on published articles. The company blamed a third-party content vendor that insists that its articles were written by humans. 

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 12/6/23

News

Google announces Gemini, an “everything machine” competitor to ChatGPT that offers little new functionality to wow users, but puts AI’s key features into a single package. The company will license Gemini to Google Cloud developers and will integrate it starting this week in Google’s consumer-facing apps such as the Bard chatbot, Gmail, and YouTube.

Business

China-based AI drug discovery company XtalPi, which has raised $732 million in funding at a valuation of $2 billion, files for an IPO on the Hong Kong exchange. The three founders are MIT quantum physicists.

The Wall Street Journal reports that drugmaker Johnson & Johnson has hired 6,000 data scientists and spent hundreds of millions of dollars on AI drug discovery technology.

Research

A study finds that ChatGPT incorrectly or incompletely answered 75% of the drug-related questions that were posed to a pharmacy school’s drug information service. It also generated fake citations when asked to list its references. The authors warn that users should check its results against trusted sources.

Other

An op-ed article in Nature says that healthcare institutions should proceed cautiously in rolling out off-the-shelf proprietary large language models from “opaque corporate interests” to avoid undermining the care, privacy, and safety of patients by using tools that are hard to evaluate or could be changed or taken offline. The authors urge a more transparent and inclusive approach in which health systems researchers, clinicians, patients, and tech companies collaborate to build open source LLMs.

The American Veterinary Medical Association describes how one member is using AI to provide a second opinion on X-ray interpretation and to record visit conversations and turn them into SOAP notes.

Microsoft research VP Peter Lee, PhD provides thoughts about AI in healthcare:

• It won’t be long before doctors will refuse to practice medicine without the assistance of AI.
• Bing and ChatGPT are good at deciphering lab test results and explanation of benefits, where they can be asked “are any of these results concerning“ or “do I owe money.”
• He and his sisters asked GPT-4 to review their father’s medical records, then list the three best things to ask the specialist in their 15-minute visit.
• He says that GPT-4 is “almost superhuman” in its ability to serve as a second set of eyes for a physician in reviewing patient information and the doctor’s diagnosis by being asked, “Did I miss anything, or should I consider something else?”
• Lee warns doctors not to think of ChatGPT as a computer that has perfect recall and performs perfect calculations, but rather as a “personal intern” whose work requires review.
• AI doesn’t follow the regulatory framework of a software medical device, so it’s up to the medical community to take control of how and when it it used.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 11/29/23

News

Amazon announces Amazon Q, a generative AI assistant for businesses that provides personalized company content to employees.

NHS England will use AI to identify patients who are at risk for an ED visit or hospital admission and could instead be contacted by health coaches, including the use of sensors on refrigerators and teakettles to identify changes in eating and drinking habits. It is also using AI to predict the  top at-risk patients who are then contacted by staff who offer social care assessments and medication reviews. NHS has also expanded its hospital-at-home program to 10,000 patients. It expects the new programs to prevent 4,500 ED visits, 17,000 overnight hospital stays, and 23,000 physician appointments.

Hong Kong’s hospital authority will pilot the use of AI to identify potentially inappropriate ordering of certain broad spectrum antibiotics following a jump in vancomycin-resistant infections. Broad-spectrum antibiotic prescribing jumped during the pandemic as COVID-19 patients used most available isolation beds. 

A Denver news outlet describes the efforts of UCHealth to apply AI to sepsis detection. CMIO C.T. Lin, MD says its initial deployment of Epic’s tool was causing alarm fatigue and a flurry of end-of-shift alerts as overworked nurses back-entered vital signs they had scrawled on their arms all day. The health system redirected the alerts to its virtual health center, where 24/7 ICU nurses monitor the vital signs of up to 500 patients and can immediately review the EHR and live video feeds. UC Health estimates that the sepsis alerts are saving 375 lives per year, with another 800 patients saved by early detection of other forms of deterioration.

Business

Drug maker Roche’s Genentech subsidiary will collaborate with microprocessor vendor Nvidia to advance drug discovery.

Vivodyne, which uses AI-powered robotics to cultivate, dose, and analyze prospective drugs on human tissues, raises $38 million in seed funding.

Research

Researchers use AI to identify brain imaging markers that may help diagnose ADHD in adolescents.

A Microsoft study finds that GPT-4’s performance on medical specialty questions can be improved by the use of more carefully structured prompt alone, without being expensively trained on more detailed data.

Republic of Korea researchers develop a machine learning model to predict cardiac arrest in ICU patients using ECG data, which they say is a better predictor than EHR information and can be used for continuous monitoring. 

Other

Sports Illustrated gets caught running AI-generated articles that are bylined to AI-generated writers with headshots that were acquired from online marketplaces. One article purportedly written by “Drew Ortiz” expertly opines that “volleyball can be a little tricky to get into, especially without an actual ball to practice with.”

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.

Healthcare AI News 11/15/23

News

The estates of two deceased men sue UnitedHealth Group, claiming that the company uses AI algorithms to deny care to seniors who are enrolled in its Medicare Advantage plans. The complaint says that the AI system that UHG gained in its 2020 acquisition of NaviHealth has a 90% error rate and creates generic care recommendations that are used to override physician assessments of medical necessity, placing the Medicare Advantage patients at a disadvantage compared to traditional Medicare. A previous report found that patients who were entitled to up to 100 days in a nursing home following hospital discharge were usually cut off within 14 days by UHG.

Elsevier and OpenEvidence create an early access version of ClinicalKey AI, which allows doctors to enter patient information that is matched against Elsevier’s medical journal data to help them make clinical decisions. OpenEvidence’s original system is available for demonstration.

OpenAI pauses new signups for ChatGPT Plus due to a user surge following the company’s November 6 DevDay. The company also acknowledges that it is working on GPT-5, which will require more public and proprietary data sets.

Business

The CEO of an AI-powered robotics company says that the factors that made ChatGPT successful – training on a vast and diverse dataset and aligning the model’s response with what humans find useful – are being applied to AI robots that can understand their surroundings, make decisions, and adapt to changing circumstances. He predicts widespread deployment of robots that can perform object manipulation in 2024.

University of Miami Health System will use Aidoc’s technology system-wide  to identify and triage abnormalities in patient scans. The Israel-based company offers 13 FDA-cleared algorithms. The CEO is a former AI leader in the Israeli Air Force and co-founded the company with military peers one year after discharge.

Montreal-based Pathway, which provides evidence-based responses to clinician queries, raises $5 million in seed funding.

Research

Researchers in Australia refer to large language models that have limited guardrails as “weapons of mass disinformation” in a JAMA Internal Medicine article, where they describe using ChatGPT to create 102 disinformation blog posts about vaccines and vaping. Creation of the posts that totaled 17,000 words, which included fake patient and clinician testimonials and scientific-looking references, took just over one hour. Additional OpenAI tools contributed realistic images and a deep-fake video.

AI outperformed tissue biopsies in grading the aggressiveness of retroperitoneal sarcoma, which could potentially spare low-risk patients from unnecessary treatments and tests.

Tech-heavy medical practice operator Forward Health develops a standalone, app-powered  medical station it calls the CarePod, which it hopes to place in malls and office buildings where members will pay $99 per month for access. Founder and former Google computer scientist Adrian Aoun, who says “I don’t even believe a doctor’s office should exist,” hopes to turn healthcare into a product rather than a service. CarePods are powered by a large language model that uses medical research data to create care plans. The company just raised a $100 million Series E round and hopes to launch 25 CarePods initially and scale to 3,200 within one year. Aoun says he hopes to enhance the machines to perform open heart surgery within the next 20 years.

Other

AI expert Mark Hinkle builds a GPT into which users can upload PDFs of their health insurance plan options during open enrollment, then ask ChatGPT specific questions and have it compare plans.

A Nebraska oncology research nurse’s lung cancer is detected by Methodist Hospital’s AI-powered LungGPS, a Medtronic system that double checks radiology images and reports for incidental nodules. She was X-rayed for pain in her left side, but LungGPS noted nodules on the other side, with the early detection allowing her stage 1 lung cancer to be cured via surgery. LungGPS caught 26 patients with undetected cancer in 2022 and 15 so far this year.

Contacts

Mr. H, Lorre, Jenn, Dr. Jayne.
Get HIStalk updates.
Send news or rumors.
Contact us.