So what's wrong? Health companion app uses AI, machine learning to ask smart questions

So what's wrong? Health companion app uses AI, machine learning to ask smart questions

(Tech Xplore)—Your eye does not feel right. Your gums don't feel right. Your arm aches but you cannot recall doing anything out of the ordinary.

You can surf for answers on the Internet but you are never quite sure if the information quality is right—and you are never quite sure if your search words that you entered are properly addressing all the information you would need.

Now there is something called Ada Personal Health Companion, which can be a useful app in getting to know more about what ails you.

A video posted earlier this month said that this personal companion uses AI and machine learning. Designed by doctors, data scientists and engineers, Ada is on a mission to deliver personalized health information.

Medical News Bulletin ran an article on the app, and said, "Ada can assess patients' symptoms, track individuals' health progress, and share data with doctors."

"At Ada's core lies a sophisticated medical reasoning engine that also supports doctors by providing earlier health information and informing clinical decision making," said the video notes.

So what's wrong? Headache. How long has this been troubling you? Multiple choice answers about duration include a don't know. Stephanie Mlot in provided an example.

"Plugin 'sore back,' for instance, and you'll be quizzed on things like extent and severity of the problem, environmental factors, and auxiliary issues. Ada then works her magic to provide a concise report, highlighting possible causes and next steps—can this be managed at home, or should I seek medical advice?"

The companion app uses AI technology to learn and grow to ask more sophisticated questions, as users engage with it in order to better track individual health, as described in Medical News Bulletin.

Ada improves with every step of symptom assessments by learning from the user's responses to each question.

Steve O'Hear in TechCrunch said two of Ada's founders—CEO Daniel Nathrath and Chief Medical Officer Dr Claire Novorol—explained the app has been six years in the making, and started life out "as being doctor-facing, helping clinicians to make better decisions."

But wait. Where is the line between self-help information and taking that bus and two trains to the doctor's office for a walk-in consultation?

Ada's team is not trying to convince people to ignore the doctor's receptionist, but to understand medical questions that one may want to explore when consulting a physician.

"Our assessments are developed purely for informative purposes in order to help people better understand their health."

In a blog, they said that "Through our app, we supply doctor-vetted content at your fingertips and algorithm-based machine learning behind the scenes."

O'Hear pinpointed the value in the app being able to help act "as a prescreen consultation before, if needed, being handed off to a real doctor for further advice, or simply helping to create a digital paper trail before a consultation takes place."

As for privacy concerns, the Ada site said that "Ada does not share your data with anyone unless you explicitly tell us to do so."

Ada has offices in Berlin and London. Daniel Nathrath, CEO, trained as a lawyer in Germany and earned his MBA from the University of Chicago. Dr. Claire Novorol, MD PhD, is chief medical officer. Dr. Martin Hirsch, PhD, chief scientific officer, has interests that include cognitive neuroscience and semantic knowledge representation.

Explore further

'Ask a Doctor' surfaces reliable information about heart conditions

More information:

© 2017 Tech Xplore

Citation: So what's wrong? Health companion app uses AI, machine learning to ask smart questions (2017, April 23) retrieved 8 August 2020 from
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Feedback to editors

User comments