Using facial recognition for medication adherence?

While doing a routine search of Twitter I came across AiCure (@AiCureTech), which touts itself as “Computer vision and facial recognition technology to confirm medication adherence on mobile devices”. Ok, you got my attention. Unfortunately the Twitter account appears to be dead as the last Tweet listed on the account was from September 25, 2013. The AiCure website is a bit more recent, however. The last item posted to their News & Events section was from March of this year.

There’s a video on the website that gives a basic overview of the process. I would have embedded the video here, but couldn’t figure out how to grab it, which is a real shame because it’s in their best interest to make information easy to share.

After watching the video I’m not entirely sure that the process makes sense to me. The video shows a jogger running on a pier. The jogger receives a notification on her smartphone reminding her to take her medicine. She stops, pops the tablet in her mouth, records the transaction via facial recognition on her smartphone, and then merrily continues on her way. In my experience people that are as “with it” as the person portrayed in the video don’t have any trouble remembering to take their meds; calendar reminders, pill bottle next to the coffee pot, etc. And why is the jogger carrying her medication with her while out jogging? I assume her jogging session wouldn’t last more than an hour or two. Take the med before or after. There’s no sense of the importance of the medication to the patient’s condition, nor is their any sense of the person being so busy that they couldn’t remember to take their medication. It would have made more sense to show some teenager with a serious medication-dependent disease state going through a busy school day. Right? Having so much fun with their friends that they forget to take their medication?

Thoughts on marketing aside, the concept of using facial recognition is intriguing.

From the AiCure website:

The combination of automated computer vision technology with dynamic patient feedback, offers a new gold standard in medication adherence monitoring. The computer vision platform is being extended to develop a robust identification and authentication system for medication.

Much like a voice recognition system, which understands what the user says, AiCure’s sophisticated, patented computer vision system visually understands what the user is doing.

The software-based technology is uploaded onto a smartphone or tablet computer. The user follows a series of pre-determined steps that are instantly recognized and confirmed through the webcam.

Automated DOT® [Directly Observed Therapy] confirms facial identity, medication dosage, correct ingestion, and time of ingestion. In addition, built-in data tools allow for ongoing patient-provider feedback; reminders in case of nonadherence; positive feedback; self-reported data by the patient; and therapy information – all designed to ensure real-time adherence monitoring and improved patient adherence over time.