6
Machine Learning Powers a Holistic
Approach to Healthcare
I thought
you would
be glad to
know that
Aidoc just
saved one
patient's
life."
Original article published by WIRED.
"
"Hey" the
doctor texted,
All three companies working with machine learning through AWS agree: This
is only the beginning. While they've accessed unprecedented levels of scale
through AWS Machine Learning, there's plenty of work still to come in applying
machine learning's success in diagnosis in other areas of healthcare.
Hamilton envisions machine learning creating more opportunities for patients
to have productive conversations with their healthcare providers. The company
is currently developing natural language processing (NLP) technology that
could free doctors from entering data into their laptops by intelligently
listening to the dialogue in the exam room; the NLP model can take notes and
pull up relevant test results or images. Background applications might surface
alternative diagnoses for the doctor based on millions of records.
And the care doesn't end when the visit is over: Back at home, the patient
might receive interactive prompts: "Did you fill your prescription? Would it
help if we delivered it?" Remote monitoring of wearable technology could
congratulate the patient for reaching 3,000 daily steps or warn case managers
that rapid water weight gain indicates pending heart failure.
The goal, Hamilton said, is "trying to get to a conversational healthcare
workflow not a visit-oriented, venue-oriented workflow."
While these possibilities continuing to drive excitement around machine
learning, companies agree that the progress to date is gratifying—especially
when it can save a life. Wallach says he's continuously motivated by the direct
messages he gets from doctors on a weekly basis—including a recent one
about an ER patient with a massive pulmonary embolism.
Ed Wallach,
Aidoc CEO
