AI in healthcare: Prediction is dead, Realization is coming.

n engineering, we primarily use feedback loops to create and improve systems.

Aeronautics is one such example. Fifty years ago, the accident rate was much higher. Today, it has drastically reduced because, with every accident, multiple sensors analyze the incident, leading to global software updates that prevent similar accidents from happening again.

On the contrary, you would never see that application in healthcare and medicine.

The lack of dynamic measuring tools has relegated the healthcare industry to a reactive rather than a predictive stance.

One would go to the doctor only if one has an issue, and the doctor would try static tools to mitigate the effects.

In more detail, one might be instructed to undergo blood analysis, which represents a static sample at a given moment.

Then, the doctor will compare those values with the population to create a diagnosis.

In engineering, statically testing a dynamic system would be laughable.

There are multiple issues here. Primarily, each individual has different baseline values for each biomarker. So, comparing with anyone else is wrong by definition.

For the last few years, we have seen the emergence of dynamic sensors.

Continue the full article here