The finessing of data.

In 2017, a group of us were in ballroom at a Hilton in Orlando, FL with the local chapter of HIMSS. On stage was a radiologist from the University of Chicago who had captured the entire audience with his straightforward and concise explanation of machine learning; the top buzz term of the year.

“Everyone pull out your smart phone. I have an iPhone so I am going to show you from Photos, but the same could apply to Google Photos as well- open up the app!” He said with a great deal of excitement from the podium.

He utilized the search function to type in things like “beach”, “drink”, “meal” to pull up vacation pictures in seconds vs. the old way of trudging through folders of pictures saved in folders on his laptop computer. Remember, it was 2017 so this was a big deal not having to defer to folders on a computer.

After he heard the wonderment and surprise in the audience, he quickly pivoted back to healthcare. “This is what machine learning does in radiology. I can quickly search for a mammogram image because the picture is always going to have similar attributes in shape, black space, and other components to make clinical research and trials… we can use patters (or machine learning).”

I think back to this talk often. Patterns- I am watching my kids learn about this in pre-school and kindergarten. They can track patterns and predict things faster than most adults because it is so fresh in their way of thinking. Patterns can help predict events, behavior, and more. In the case of clinical workflows, and other research arenas, patterns and prediction (with rigor and logic) can save time to diagnose.?With the workforce shortage and pent up demand for healthcare alone, this interpretation and synthesis of data is (still) so exciting.

I am referencing an event from five years ago, which in the world of health IT and digital health feels like a lifetime ago. It still resonates though… why? The growth of data interpretation and utilization is still being finessed, so without mastery there still lies questions in the greater market. All the while, many still don’t fully grasp the concept of how this helps society at large. Why should buying behavior in the consumer world be paired with clinical behavior by a physician, health system, or health plan? Why are we still just scratching the surface of the impact of broader health- financial and physical health? And then there is privacy. So, can we keep these exercises in data interpretation and machine learning anonymized at all times and should we?

I would love to hear from you.