With the help of AI, smartphones can detect signs of knee problems years before patients feel any pain, Dr. Florian Hinterwimmer, Technical AI Lead of the AI in Orthopaedics (AIO) research group at the Technical University of Munich, explains in his DLD Health talk.
Statistically, joint decay starts when people are older than 40. “Micro-shifts start to happen”, Hinterwimmer explains. “Not much, it’s asymptomatic, you don’t really feel anything, but things are slowly changing.”
Around 600 million people worldwide are affected by osteoarthritis, a degenerative joint condition that makes life painful and miserable for many patients, Hinterwimmer notes. “It hurts. You cannot go skiing anymore. You cannot play with your grandchildren, whatever you like to do.” At the same time, “we have clinicians, mostly orthopedic surgeons, which are overburdened, overwhelmed.”
To address this problem, Hinterwimmer’s research group develops an app called NewKnee that continuously collects movement data, such as walking speed and walk asymmetry, directly from smartphones without extra hardware. An AI model can then flag health risks early and predict joint deterioration up to five years before a clinic detects it.
“We’re always collecting symptomatic data, when the patient’s already sick”, Hinterwimmer says. “With this approach, we can collect lots of data before the patient is sick – or at least when the patient’s sickness is at a lower level.”
Watch the video to explore this talk in detail.
