We humans learn from experience (I hope!)… 🤞
More quality experiences and data points we have, the more we improve our understanding of a task, subject, skill… 🎹 🤹♂️ 🎭 🧮 📚 ⛹️♀️
Same can said with machine learning, especially when used as a (or within a) medical device.
As the FDA puts it: “One of the greatest potential benefits of ML resides in its ability to create new and important insights from the vast amount of data generated during the delivery of health care every day.”
They’ve approved over 500 AI/ML-enabled medical devices for use in the US (I put a link to the list in the comments). 👩⚕️ 🏥
They acknowledge that the traditional medical device regulations were not designed for adaptive AI and ML technologies. They do not allow for adaptive learning—where the software learns from an ever-expanding data set—without additional premarket approvals, which is a problem.
So how then to change the regulations? What kind of framework should be used to regulate AI/ML as a medical device? 🤔
This is a critical question.
It’s also the topic of the panel I’ll be moderating in Malta at the Med-Tech World event Thursday morning. 🇲🇹
Joining me on the panel are:
Dr Hugh Harvey
Daniel Torres Gonçalves
I’m looking forward to hosting a very interesting discussion.