Detecting Patterns
This is how I imagine a future visit to a radiology department to be when AIs do the diagnostics.
I’ll walk into the radiology section and will be instructed to go to the booth. A disembodied voice will tell me where to stand and to position myself.
After the scan that same voice will read out my diagnosis. If it’s not programmed to understand my question I will battle to find a person to talk to. The only person who can really answer my question is the radiologist-turned-engineer halfway around the world.
He would say something like: based on pattern match of the size, shape and position of the blob found in your body you are likely to be suffering from the diagnosed condition. This was cross-referenced with the database of other humans who share my correlated genetic markers.
A treatment plan might be generated or a referral will be made. If it turns out that the diagnosis was incorrect based on my response to the treatment then additional data points based on my profile will be ingested by the pattern recognition and deep learning network. This will improve the diagnostic and predictive capability of the machine.
Zaheer Ismail
Comments