Follow Datanami:
September 25, 2020

COVID-19 Symptom Evaluation App, Built with Watson, Becomes Call for Code Regional Finalist

As the COVID-19 pandemic has continued to spread, doctors and hospitals have been overwhelmed by anxious patients worried about a symptom and unclear on what COVID-19’s symptoms are or which warrant a visit to a medical professional. Altran, a design and engineering services firm and subsidiary of French multinational Capgemini, used IBM Watson to create an intuitive, AI-driven app for users to evaluate their symptoms and help them to decide next steps.

The app’s interface. Image courtesy of Altran.

The app, called “Are You Well?”, is built on a “scalable, extensible, resilient multi-sided platform” using IBM Watson Assistant, IBM Discovery Service, IBM Cloudant DB and other IBM services. It first asks users to enter demographic info. Then, based on that demographic information, the Watson-powered AI asks the user a series of questions: what is your current body temperature, if you know it? Where have you traveled? Are you experiencing a running nose, sore throat, headache, dry cough, or weakness? As users answer questions, a series of parameters defined by health experts gradually combine to form a risk assessment for each user: high, medium, or low. 

If the user has worrying symptoms, the app recommends self-isolation and requests further health data, even offering a progress bar for a given user’s quarantine. In the event of a “low” risk assessment, the app offers a series of core recommendations: maintain social distancing, stay home, wash your hands, and so on. Altran, of course, reminds users that the app is not a substitute for genuine medical advice: “Follow a doctor’s advice at all times,” the introduction video says. “Only he will know what to do.” 

To that end, doctors or medical centers integrated with the app can access a “global view” of patients’ color-coded survey outcomes. Zeroing in on an individual patient allows the doctor to see detailed travel and quarantine history, risk level, symptoms, demographic information, body temperature graphs, and even a chat box to talk with the patient. Administrators can configure the questionnaire and even tailor the algorithm for their specific implementation.

IBM has selected the app as one of its regional finalists for the Asia Pacific region in its 2020 Call for Code Global Challenge. The Call for Code Global Challenge, which previously focused solely on climate change, this year allowed applicants tackling COVID-19 as well. The winners – who will be chosen by a panel of esteemed judges including Bill Clinton and Mark Cuban – will receive prizes ranging from $10,000 to $200,000, with the grand prize winner also receiving help in implementing their solution at scale.

Related Items

New Analytics Tools Predict COVID-19 Patient Mortality

Getting Creative with COVID-19 Contact Tracing

Sandia Labs Model Predicts COVID-19 Cases Over a Week into the Future