It seems like AI is everywhere this Summer. But what does it mean in the world of healthcare? How is it actually applied? Check out Ambra Health CEO, Morris Panner, discuss on Bloomberg Radio.
You can also read more about AI and particularly, how data has moved from a frustrating liability to a critical piece of the healthcare treatment puzzle in this article by Morris that originally appeared on June 12, 2017 in HIE Answers.
Artificial intelligence (AI) is the future of medicine. Facebook uses AI to suggest friends’ faces to tag in a photo; Spotify to suggest your perfect playlist. Healthcare is far behind, but catching up.
Yet, none of this will be possible unless we properly manage our medical data, just as consumer apps have corralled our personal data. Our own medical studies, our pathology results, our CAT scans, and our lab values enable this medical revolution. This astonishing transformation in how we think about healthcare data poses technical and ethical challenges galore.
Imagine you were a miner during the first gold rush. As tools got better, you could dig deeper. Your claim was worth more. The same is true for algorithmic insight – as it sharpens, data itself becomes more valuable.
Sometimes it can take a long time for data to be useful – we just don’t understand what we see. A CAT scan today may mean nothing, but ten years from now, hidden in the pixels, may be cures. Only a computer can manage this vast data context challenge through space and time.
Historical healthcare data used to be known as: “STUFF THE GOVERNMENT MADE YOU KEEP,” “OLD DATA SOMEONE COULD USE TO SUE YOU WITH,” even “ALL THE UNNECESSARY MEDICAL STUDIES NO ONE EVER REALLY LOOKED AT.”
As a result, most institutions treated data storage as a burden. Data was a liability. Things have changed. We want our data to live forever.
Data is not created equal. Your doctor’s notes are interesting, but subjective. A recent study found an uncomfortably high percentage of physicians reviewing their own findings (scrambled to hide identity), ended up disagreeing with themselves.
Imaging, pathology studies and lab values, on the other hand, are known as Immutable Data, the true objective map of human disease or its precursor. Source of truth, digital data a is the most important “claim” in the gold rush to mine information to cure disease. The meaning of Immutable Data may be obscured for decades, but the information it contains never changes.
To enable breakthroughs, we must appropriately store, curate and share Immutable Data. Last generation systems fall short. Data size and complexity makes storage difficult. Imagine storing millions of HD movies for decades and hoping none of the pixels ever changed. Then, imagine trying to share and manipulate all that data. One hospital might have a treasure trove of X-Rays, but be unable to share them with a partner hospital.
The devil truly is in the technical details, as the PwC Center for technology and innovation points out.
With the data revolution in the consumer space, we never expect to pay for Facebook or Google. Data has enabled services to be provided for free — we finance it with our information. As the saying goes, we are the product.
The impact of healthcare data will be different, but no less revolutionary. Federal privacy rules, known as HIPAA, protect patient data from disclosure and inappropriate use. In general, you legally own your own health data. Yet, there is an exception for research with anonymized data.
Will hospitals or data storage companies end up owning this valuable data? Since our information is the key asset for research should what we learn be public property? Should we make it easier for researchers to acquire data, so cures can be developed more rapidly? Never before asked questions abound!
Immutable Data – our data – is the key scientific asset of our time. How we enable our miners will set the stage for advancing health for years to come.