Blog
Subscribe to my Substack
Accessible, AI-powered Health Tracking: More Data, Less Clarity?
Why the Rise of Advanced Health Monitoring Demands a Renewed Focus on Statistical Literacy
With the increasing amount of data from wearable devices, blood tests, and genetic sequencing combined with artificial intelligence (AI), it’s never been easier or more inexpensive to keep track of your health and catch potential diseases early. But we’re leaving something crucial behind: statistical literacy. If we can’t correctly interpret and take the best actions with the information we’re getting,then what good is it?
I discuss how, due to poor statistical literacy, accessible health informatics through blood tests, wearables, and genetic tests could cause more unnecessary procedures, anxiety, and costs without getting much in return. To mitigate this issue, we need to place a greater emphasis on how AI-driven health information is presented and in which scenarios it should be provided.
One Big Reason Why Artificial Intelligence Isn’t Disrupting Healthcare: Paying for It Is Too Difficult
Traditional reimbursement models stifle innovation and support the status quo by raising development costs and delaying time to market
So far, AI/ML has had a limited impact on the healthcare industry. The truth is, disrupting healthcare with AI/ML is going to take more than just cutting-edge technology: paying for it will need to entirely change, too.
Companies developing AI/ML health products currently rely on traditional reimbursement systems to make sure their products can reach patients. But by doing so, AI/ML development encompasses higher costs, delayed times to market, and larger barriers to entry. Most of all, it prevents smaller companies from tackling large problems in innovative ways.
In this post, I’ll explain the problems with the current reimbursement paradigm for AI/ML development and the ways we can fix it by taking a page out of the enterprise software playbook.
Leveraging Machine Learning Using Digital Twins in Alzheimer’s Disease Clinical Research and Beyond
Three Practical Applications Surrounding the Intersection of Machine Learning, Precision Medicine, and Clinical Trials
UC Irvine Institute for Memory Impairments and Neurological Disorders
Precision medicine aims to target the right treatments to the right patients at the right time. By leveraging big health data and machine learning, we can achieve personalized treatment for every patient through what is called “digital twins”. The adoption of digital twins could greatly impact clinical research by saving clinical trial costs, accelerating drug approvals, and optimizing the use of available treatments. In this blog post, I discuss how digital twins can be used in Alzheimer’s disease research and cover the opportunities and challenges surrounding three use cases: minimizing control arm subjects, trial enrichment, and precision medicine.
Can Machine Learning Give Us Faster and Cheaper Clinical Trials?
On Trial Design, Historical Controls, and Clinical Prediction Models
In drug development, where the median clinical trial costs about $48 million, biotech companies face the major challenge of balancing expenses with the need for robust, well-designed, and sufficiently large clinical trials. This reality has motivated software companies like Unlearn.AI to pioneer machine learning (ML) applications to boost clinical trial efficiency. Yet, implementing such cutting-edge methodology is not without significant practical challenges.
Today’s post delves into the merits and limitations of Unlearn’s ML-based Prognostic Covariate Adjustment (PROCOVA). I argue that the theoretical cost-savings of PROCOVA are unlikely to be realized in practice. Given the often high stakes of clinical trials, the risk of adopting innovative ML-based methodologies may not be worth the purported benefits.
The Hard Truth about Artificial Intelligence in Healthcare: Clinical Effectiveness is Everything, not Flashy Tech
Lessons from HeartFlow’s Incursion into Cardiovascular Imaging
AI in healthcare faces a critical issue: our obsession with cutting-edge technology often overshadows the actual impact on patients. Successfully bringing AI medical devices to market entails much more than excellent diagnostic performance; it requires rigorous clinical trials and comprehensive cost-effectiveness analyses. HeartFlow’s AI-powered cardiac imaging product FFRCT is a perfect example of that. In this blog post, I critically review FFRCT and discuss broad lessons for the future of AI medical devices