Sickness detection. Google's AI approach. A-ha! keynote
Google, Can I Stay Home From School?
When I was a kid, I could never pretend to be sick in front of my parents. They’re doctors, and just by looking at me, they could tell if I was lying. If I was actually sick, they also knew before I even told them. I never understood how.
Turns out, there are real visual cues that can help us to detect when a person is sick based on changes in skin color, shape of the eyes, and mouth, according to a new paper. The findings show that even non-doctors can use these cues to help guess when someone is sick.
As machine vision gets better, is it possible that computers will also be able to use these cues (and more) to tell if you’re sick?
This is the path Google is going down with a new patent we caught and analyzed. The patent talks about using an optical sensor in your bathroom to detect any blood flow abnormalities. This is intended to help patients at risk for cardiovascular disease.
Google is one of several tech giants ramping up their patent activity in the healthcare space, which we’ve analyzed before.
Google’s AI Trident
This patent falls into Google’s three-pronged strategy with AI. The company is using AI to:
1) Replicate the efficacy of a doctor for specific tests
2) Organize massive amounts of information to support doctors
3) Use AI to do things humans could never do, like analyze huge datasets
The first bucket is clear in its patents. Along with this new one, we talked about two others Google has for an abdominojugular reflux test and a concussion test in our recent “consumerization of healthcare” research briefing.
Bucket 2 falls closer to its DeepMind division, which is working with the NHS to help do risk assessment and triaging within hospitals as one of many projects.
And the last bucket falls closer into Google’s other divisions, including Verily and spin-out Calico. Both of these companies use AI to attack massive datasets that are too large for humans to navigate. Verily does this from a population health lens, while Calico is doing this to understand why we age. We spoke to Daphne Koller, Chief Computing Officer of Calico, on how the company operates across these spaces.
Bop It, Pull It, Buy It, Flip It
If you thought quick gains were only a cryptocurrency thing, you must have missed the Impact Biomedicines story. The company raised its first institutional financing from Medixci Ventures, a $22.5M Series A, in October. Less than 3 months later it sold to Celgene for $1B+ in cash and up to $7B in payouts.
This is another cancer bet for Celgene, but falls in line with its increased activity in the cancer therapeutics space.
You can see our full collection of 1100+ cancer therapeutics here.
Can I speak with your boss?
If I haven’t convinced you that new technologies present a real threat that healthcare corporates should care about, maybe our CEO Anand can. We released our results from our “State of Innovation” survey to help understand why and how you should think about innovation. Here’s his keynote speech.
