AI Rx: Your Weekly Prescription for News on AI in Healthcare

This week AI is: fighting insurance claim denials, personalizing addiction recovery, reshaping medical education and using Tom Hanks’ likeness to promote fake “miracle cures and wonder drugs.”

Aug 30, 2024

Hi! If you’re interested in AI and healthcare, you’ve come to the right place. This newsletter is a collection of articles, studies, op-eds and other news from the past week involving how AI is being used in healthcare. As a former biotech reporter and current communications consultant to the biopharma industry, I’m fascinated by how fast AI is being adopted across all industries, and its potential to improve healthcare. Check out the stories below and be sure to subscribe and share on your social networks!

Ryan Flinn

Tom Hanks Battles AI Doppelgangers in Health Ad Scam

What's the News: Tom Hanks took to social media to warn about AI-generated ads using his likeness to promote "miracle cures and wonder drugs" without his consent. The Oscar winner emphasized that he only works with his board-certified doctor for his type 2 diabetes treatment.

So What: This incident highlights the growing concern over AI's ability to create convincing deepfakes, potentially misleading consumers and eroding trust in public figures. As Hanks pointed out, "Anybody can now recreate themselves at any age they are by way of AI or deep fake technology."

To Be Sure: While AI technology offers exciting possibilities in entertainment and beyond, this case underscores the need for robust regulations and ethical guidelines to protect individuals' digital identities and prevent fraudulent use of AI-generated content.

Read More: Tom Hanks Warns Of AI-Made Drug Ads Using His Likeness: "Do Not Be Swindled", in Deadline, by Glenn Garner

Kids' Hospital Embraces AI: 5 Projects That Could Change Pediatric Care

What's the News: Washington, D.C.'s Children's National Hospital is launching five artificial intelligence projects in collaboration with Virginia Tech, aiming to transform pediatric healthcare. The initiatives range from predicting emergency departmant surges to generating facial images for better detection of genetic syndromes.

So What: These projects could significantly improve pediatric care, especially for children with rare conditions. As Dr. Marius George Linguraru of Children's National puts it, "AI's potential to offer life-changing solutions for children with rare medical conditions is immense."

To Be Sure: While the potential benefits are exciting, implementing AI in pediatric care requires careful consideration of ethical implications and data privacy concerns, especially when dealing with sensitive information about minors.

Read More: 5 AI projects move forward at children's hospital, in Health IT by Giles Bruce

Uncle Sam Wants... To Monitor AI-Enabled Medical Devices

What's the News: The U.S. Department of Health and Human Services is funding a new initiative called PRECISE-AI to develop tools that can automatically monitor and maintain the performance of AI-enabled medical devices.

So What: With over 950 FDA-authorized medical devices now integrating AI, ensuring their ongoing reliability is crucial. This initiative could lead to more trustworthy AI tools in healthcare, potentially improving patient outcomes and clinician confidence in using these technologies.

To Be Sure: While this project addresses a critical need, the complexity of healthcare settings and the rapid evolution of AI technology mean that maintaining device performance will likely remain an ongoing challenge requiring continuous innovation and oversight.

Read More: HHS to fund AI-enabled medical device maintenance tools, in Healthcare IT News, by Andrea Fox

The Identity Crisis Holding Back Healthcare AI

What's the News: A recent survey of hospital, health system, and payer executives reveals that 57 percent believe errors in patient data matching could lead to a healthcare crisis in the next 5-10 years. Nearly half point to data fragmentation across systems as a major challenge.

So What: Accurate patient matching is crucial for the success of AI in healthcare. As Avi Mukherjee and Andy Dé of Verato explain, "When lives are on the line, clinicians cannot rely on 70 percent accuracy. They need absolute confidence that the information they see pertains to the correct patient — every time."

To Be Sure: While AI holds promise for improving healthcare efficiency and outcomes, fundamental data management issues must be addressed first. Without solving the patient identity puzzle, even the most advanced AI systems may struggle to deliver reliable results in real-world healthcare settings.

Read More: The Importance of Accurate Patient Matching for AI Projects, in HIT Consultant, by Avi Mukherjee and Andy Dé

AI's 'Hallucinations': A Pharmaceutical Nightmare?

What's the News: As the pharmaceutical industry embraces AI, machine learning, and large language models to process vast amounts of information, concerns are growing about the technology's tendency to sometimes make mistakes or 'hallucinate' false information.

So What: These AI 'hallucinations' could have serious implications in drug development and patient care. The industry is now looking at new technologies to rebuild trust and ensure the reliability of AI-generated insights in pharmaceutical research and development.

To Be Sure: While AI holds immense potential to accelerate drug discovery and improve healthcare outcomes, its integration into the pharmaceutical industry must be approached with caution. Ensuring the accuracy and reliability of AI-generated information remains a critical challenge.

Read More: As pharma's AI revolution gets underway, 'hallucinations' pose a great risk, in Pharma Voice by Michael Gibney

Ethics: The Secret Ingredient in Pharma's AI Recipe

What's the News: The pharmaceutical industry is rapidly adopting AI, but concerns are rising about potential ethical pitfalls. Industry experts are calling for a holistic approach that prioritizes ethics in AI development and deployment.

So What: Ethical AI use in pharmaceuticals could revolutionize drug development, optimize pricing models, and improve access to life-saving treatments. However, without proper safeguards, AI could exacerbate health disparities, compromise patient privacy, or be manipulated for profit over patient well-being.

To Be Sure: While AI offers exciting possibilities, the industry must prioritize transparency, fairness, and patient well-being over mere efficiency or profit maximization. Robust governance frameworks and a culture of ethical stewardship will be crucial in navigating the AI revolution in pharmaceuticals.

Read More: In the Pharmaceutical Industry, Ethics are Still More Valuable Than AI, in MedCity News by Ron Tilles

Cybersecurity and Equity: Healthcare's AI Gatekeepers

What's the News: As healthcare institutions explore AI's potential to improve patient care and research, experts warn that addressing cybersecurity vulnerabilities and ensuring equitable access must be top priorities before widespread AI deployment.

So What: Robust cybersecurity measures are crucial to protect sensitive patient data and prevent disruptions to critical healthcare services. Meanwhile, thoughtful AI implementation could help reduce health disparities, but poorly designed systems could exacerbate existing inequalities.

To Be Sure: While AI offers promising solutions to many healthcare challenges, its successful integration requires a coordinated effort from regulatory bodies, the private sector, and the public. Modernizing regulations like HIPAA and creating industry-wide cybersecurity standards will be key steps in this process.

Read More: Deploying AI? Healthcare Leaders Must First Consider Cybersecurity and Equity, in Healthcare IT Today by Tim Boltz

Mayo Clinic's AI Army: 200 Algorithms and Counting

What's the News: Mayo Clinic is currently developing over 200 AI algorithms, including an advanced AI-bioinformatic system called ophthalmology parametric universal search (OPUS). This system allows Mayo to identify specific patient cohorts within its medical records, crucial for creating databases to train AI.

So What: As Dr. Raymond Iezzi from Mayo Clinic explains, "By curating annotated datasets, we can better find patterns of disease and assemble cohorts of patients for research." This massive AI initiative could significantly accelerate medical research and improve patient care.

To Be Sure: While the development of such a large number of algorithms is impressive, their effectiveness and impact on patient outcomes remain to be seen. Ensuring these algorithms are unbiased, accurate, and clinically validated will be crucial for their successful implementation in healthcare settings.

Read More: Mayo Clinic building 200 algorithms, in Health IT by Naomi Diaz

Patients Warm Up to AI in Healthcare, Survey Shows

What's the News: A survey by Talkdesk reveals growing optimism among consumers about AI's potential to improve healthcare experiences. Half of U.S. patients believe AI will enhance their patient experience within the next year, with men and millennials showing particular enthusiasm.

So What: Patients are eager for AI to automate administrative tasks like scheduling appointments and managing medication refills. This could streamline healthcare processes and reduce barriers to care, especially for stigmatized health issues.

To Be Sure: While patients are open to AI for administrative tasks, they still prefer human interaction for medical advice and discussing personal health information. Concerns about AI accuracy, data privacy, and the potential loss of human touch in healthcare remain significant hurdles.

Read More: AI Holds Promise for Improving Patient Experience, Survey Finds, in HIT Consultant by Fred Pennic

AI: Your New Addiction Recovery Buddy?

What's the News: Recent research supports the use of generative AI as a tool in overcoming addictions, both substance-related and behavioral. AI can provide personalized support, offer coping strategies, and even assist in motivational interviewing techniques.

So What: AI's 24/7 availability and personalized approach could provide crucial support between therapy sessions or when immediate help is needed. It could also help reduce the stigma associated with seeking help for addictions.

To Be Sure: While promising, generative AI is not a silver bullet for addiction treatment. It should be part of a comprehensive treatment plan overseen by healthcare professionals. There are also concerns about AI potentially giving harmful advice or perpetuating biases, highlighting the need for careful implementation and oversight.

Read More: Overcoming Addictions Via The Powers Of Generative AI, in Forbes by Lance Eliot

AI in Medicine: Who's Checking the Algorithms?

What’s the News: Hundreds of medical algorithms have been approved based on limited clinical data, raising questions about who should test these tools and how. As AI systems become more prevalent in healthcare, scientists are debating the best approaches for validating their effectiveness and safety.

So What: Proper testing of AI systems in medical settings is crucial but complex. While some hospitals are eager to adopt new technologies, others are more cautious, recognizing that regulatory approval doesn't guarantee a device's benefits. This has led to debates about the need for additional testing and validation of AI tools in real-world clinical settings.

To Be Sure: The success of AI in medicine isn't just about technical precision. Human factors play a crucial role, including how healthcare professionals interact with and interpret AI recommendations. There's also the challenge of obtaining informed consent from patients for AI use, especially in emergency situations.

Read More: How do you test AI in medicine? In Nature, by Mariana Lenharo

AI Tools Show Promise in Reducing Health Care Inequities

What's the News: A review of emerging AI tools in healthcare suggests that these technologies could help address many of the root causes of health inequities. From identifying at-risk patients to overcoming communication barriers and reducing the impact of human biases, AI is showing potential to democratize access to quality healthcare.

So What: AI could help health systems deliver better care to underserved populations by leveraging multiple data types to predict and intervene at all stages of a patient's journey. For example, AI tools can identify subtle risk factors in patient records, provide real-time translation services, and even help reduce unconscious biases in pain assessment and disease diagnosis.

To Be Sure: While AI holds promise, its implementation must be approached carefully to ensure it doesn't exacerbate existing inequalities. Concerns about data bias and the need for diverse representation in AI development and clinical trials remain significant challenges to overcome.

Read More: How AI Could Help Reduce Inequities in Health Care, in Harvard Business Review by Carol Cruickshank

AI Goes to Med School: Reshaping Medical Education

What's the News: AI is rapidly changing medical education, offering personalized learning experiences and supporting educators. AI-powered tools are being used for everything from adaptive learning platforms to virtual patient simulations and automated assessment.

So What: These AI innovations could potentially enhance the quality and efficiency of medical education, helping to produce better-prepared healthcare professionals. For instance, AI can help identify at-risk students, reduce administrative burdens on educators, and provide students with more realistic training environments.

To Be Sure: The integration of AI in medical education also presents challenges, including ensuring equitable access to AI tools, developing ethical governance frameworks, and balancing AI use with the development of human expertise and empathy. Educators stress that AI should enhance, not replace, human judgment in medical training.

Read More: How AI innovation is reinventing medical education, in Health IT by Dr. Rubin Pillay

Your Voice Could Be Your New Glucose Monitor

What's the News: A study published by Klick Labs has confirmed a link between blood glucose levels and voice pitch. The research, involving 505 participants across different glycemic statuses, found that an increase in glucose levels corresponded to an increase in the fundamental frequency of the voice.

So What: This discovery opens the door to potential non-invasive glucose monitoring methods for people with type 2 diabetes. Voice-based glucose monitoring could be as simple as speaking into a smartphone, potentially revolutionizing diabetes management for millions worldwide.

To Be Sure: While promising, this research is still in early stages. Significant further study and development would be needed before voice-based glucose monitoring could become a practical reality. Issues of accuracy, reliability, and individual voice variations would need to be addressed.

Read More: New Diabetes Research in Scientific Reports Links Blood Glucose Levels and Voice, in Financial Post (Business Wire)

AI Decodes Autism's Genetic Fingerprint in Brain Scans

What's the News: A multi-university team led by University of Virginia professor Gustavo K. Rohde has created an AI system that can identify genetic markers of autism in brain images with 89-95% accuracy. The system uses a novel technique called transport-based morphometry (TBM) to analyze brain structure patterns.

So What: This breakthrough could lead to earlier and more accurate autism diagnosis, potentially allowing for earlier interventions. The ability to link genetic variations to brain structure could significantly advance our understanding of autism's biological basis.

To Be Sure: While promising, this research is still in its early stages. Implementing this technology in clinical settings would require extensive further testing and validation. There are also ethical considerations regarding genetic screening and early diagnosis of neurodevelopmental conditions.

Read More: UVA research cracks the autism code, making the neurodivergent brain visible, in BIOENGINEER.ORG

AI Joins the Fight Against Health Insurance Claim Denials

What's the News: Holden Karau, a former engineer at major tech companies, has developed a free AI website called Fight Health Insurance (FHI) to help people appeal health insurance claim denials. The AI generates appeal letters based on scanned denial letters.

So What: This tool could help more people successfully appeal insurance claim denials, as studies show that most denials can be overturned on appeal, but very few people actually appeal. It could potentially save users significant time and stress in dealing with insurance companies.

To Be Sure: While the AI can generate appeal letters, users should carefully review and potentially edit the output before submitting. There's no guarantee that AI-generated appeals will be more successful than self-written ones, and the effectiveness of this tool remains to be seen on a large scale.

Read More: This free site uses AI to help you fight health insurance claim denials, in BGR by Chris Smith

FDA's AI Approval Process Under Scrutiny

What's the News: A study published in Nature Medicine found that about 43 percent of FDA-approved AI medical devices lacked published clinical validation data. Many devices were not tested on real patient data, with some using computer-generated "phantom images" instead.

So What: This raises concerns about the effectiveness and safety of AI medical devices in real-world clinical settings. The researchers are calling for clearer FDA guidelines on clinical validation studies and increased transparency from device manufacturers.

To Be Sure: While this study highlights important gaps in the approval process, it's worth noting that AI medical devices still undergo rigorous testing. The FDA is likely to address these concerns in future guidance updates.

Ambient AI Gains Traction in Healthcare, Promising to Reduce Clinician Burnout

What's the News: Ambient AI, which operates in the background without active user engagement, is becoming increasingly popular in healthcare. Examples include smart wearables for patient monitoring and ambient dictation technology for clinical documentation.

So What: These technologies could significantly reduce administrative burdens on healthcare providers, potentially addressing issues like clinician burnout. For patients, ambient AI could offer more seamless and personalized healthcare experiences.

To Be Sure: While promising, the widespread adoption of ambient AI in healthcare faces challenges, including high costs and potential privacy concerns. There's also a risk of creating a divide between healthcare systems that can afford these technologies and those that cannot.

Read More: Ambient AI Is Having Its 'Moment' In Healthcare, in Forbes by Dr. Sai Balasubramanian

Health Systems Struggle with High Costs of AI-Powered Clinical Tools

What's the News: While many large health systems are adopting AI-powered ambient clinical documentation tools, the high costs associated with these technologies are creating a divide in the healthcare industry. Smaller or less-resourced hospitals may be left behind.

So What: This technology divide could exacerbate existing healthcare inequalities, with well-funded health systems able to provide more efficient care while others struggle with outdated methods.

To Be Sure: As the technology matures, costs may decrease, making these tools more accessible. In the meantime, healthcare leaders are calling for more affordable pricing models to ensure wider adoption of these potentially transformative technologies.

Read More: Cost concerns loom as health systems ramp up AI, in Health IT by Naomi Diaz

AI in Medicine: Who's Checking the Algorithms?

Read More: How do you test AI in medicine? In Nature, by Mariana Lenharo

—

{EDITOR’S NOTE - THE LINK FOR THIS STORY BELOW IS BROKEN, AND I CAN’T FIND OTHER OUTLETS REPORTING THIS, SO IT MAY BE FAKE NEWS}

~~Teen Prodigy Takes On Alzheimer's with AI~~

~~What's the News~~: Nag Shivani Puram, a 12-year-old high school junior, has developed an AI project aimed at early detection of Alzheimer's disease and brain tumors. Her model combines Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) to analyze MRI scans and detect subtle changes in brain structure over time.

~~So What~~: Early detection of neurodegenerative diseases and brain tumors is crucial for better treatment outcomes. Shivani's AI model could potentially identify these conditions before they become symptomatic, allowing for earlier interventions that could slow disease progression or even save lives.

~~To Be Sure~~: While the project shows promise, it's still in early stages and would need rigorous clinical testing before it could be implemented in real-world medical settings. The integration of AI tools into healthcare also raises important questions about data privacy and the balance between AI and human expertise in medical diagnosis.

~~Read More~~: ~~Highschooler Develops AI to Revolutionize Early Detection of Alzheimer's and Brain Tumors, in TechBullion by Uzair Hasan~~

Thanks for reading! I publish this newsletter weekly, so subscribe to get it automatically in your inbox. Please send me any suggestions, stories or ideas you’d like to see featured in the future.

Ryan Flinn

In Like Flinn Communications

ryan@inlikeflinncommunications.com

AI Rx: Your Weekly Prescription for AI Healthcare News

Discussion about this post