created: Mon 26 December 2016; status: published;
Ambient healthcare is a strategy of triggering healthcare events much earlier than it would otherwise occur. It works by using the wealth of sensors around us and machine-learning algorithms us to detect trends before a human consciously would notice.
- While tracking steps and motion, the software notices a very slight but increasing change in gait over period of a year. It decides to mention it to the user and, because of the person's age, suggests seeing a doctor to head off a future knee replacement.
- Laptop camera notices minute changes in a person's face (puffy cheeks, different skin reflectivity, redness below nostrils) and detects the person is sniffling and coughing. It suggests the user has a cold or the flu and should not go into the office. It offers to notify their workplace and schedule a video visit with their doctor.
- As person walks into work, software notices they have 5 minutes more than usual before their meeting and suggests taking the steps to get some exercise.
- Microphone data from phone notices an occasional but persistent cough lasting for weeks, indicating developing lung trouble.
- As person pulls into McDonald's, the software mentions they've eaten there 9 times in the last two months. It checks current incentives from their health insurance and finds the insurer will offer a $25 gift card for reducing that to one fast food meal per month.
- Person has been complaining off and on about shoulder pain. Software suggests the user could be headed for rotator cuff injury and should meet with a physical trainer to understand what's going on, and maybe reduce their weekly tennis games.
Question 1: What can you diagnose from individual physical symptoms?
Question 2: What can you additionally diagnose from multiple physical symptoms?