(Subscribe to our Today's Cache newsletter for a quick snapshot of top 5 tech stories. Click here to subscribe for free.)
Several people around the world self-administer insulin dose for diabetes and inhaler for asthma daily. A new wireless system, developed by a team of researchers at the Massachusetts Institute of Technology (MIT), says it can flag potential errors when patients use an inhaler or an insulin pen.
The researchers have detailed their work in an article titled ‘Assessment of medication self-administration using artificial intelligence,’ published in the journal Nature Medicine.
“Insulin pens require priming to make sure there are no air bubbles inside. And after injection, you have to hold for 10 seconds,” Mingmin Zhao, study’s author and researcher at MIT, said in a release. “All those little steps are necessary to properly deliver the drug to its active site.”
Each step also presents an opportunity for errors, and patients may not even realise when they make a mistake, MIT said.
The system uses a combination of wireless sensing technology and artificial intelligence (AI) to assess patients. A wall-mounted device, equipped with a sensor, emits radio waves that tracks a patient’s movement. And when the patient moves, these radio waves are modified and reflected back to the sensor. This data is interpreted by AI for signs of the patient’s self-administration.
The researchers developed a neural network and trained it to learn patterns indicating the use of an inhaler or insulin pen, and through repetition and reinforcement, the network successfully detected 96% of insulin pen administrations and 99% of inhaler uses, MIT said.
“Each unique movement yields a corresponding pattern of modulated radio waves that the device can decode.”
Every proper medicine administration follows a similar sequence — picking up the insulin pen, priming it, and injecting. So, the system can flag anomalies in any particular step, it added.
The device can be placed in the background at home, “like a Wi-Fi router” to track movement within a radius of 10-meters, according to MIT.
“With this type of sensing technology at home, we could detect issues early on, so the person can see a doctor before the problem is exacerbated,” Zhao noted.
The system could be adapted to medications beyond inhalers and insulin pens — all it would take is retraining the neural network to recognise the appropriate sequence of movements, he further added.