6 ms·

You’re underselling it. Measuring blood at millimoles per liter precision through skin on a watch sized device is Theranos-esque physically impossible product-m

You’re underselling it. Measuring blood at millimoles per liter precision through skin on a watch sized device is Theranos-esque physically impossible product-market fit. Tricoder level of science fiction.

copperx 4y ago
I wonder whether there are any shortcuts like ingesting a substance to make the glucose visible by the sensor.
- Cshelton 4y ago
  Bingo. Noninvasive, with any sort of accuracy that can be used for anything, is most likely impossible. So back to "invasive"... what can we do... embedding a device under your skin, right around where your watch goes, that can be powered and communicate with the watch. While also making sure the device is not rejected by the body and stays in place.. AND has no other adverse side-effects. I think this is where it'll ultimately end up for the people who want it. sub 10 min procedure to put it in and should last for quite some time (maybe dissolves after 5 years for upgrades?)
  - blendo 4y ago
    Add heart rate, blood pressure, and blood oxygen level. Have the Apple Watch call 911 if things go pear-shaped. Profit.
    - Nekhrimah 4y ago
      > Profit. Well yes, but also substantial life saving interventions become possible.
    - Spooky23 4y ago
      They should profit. That will save millions of lives.
  - MengerSponge 4y ago
    Eversense sells something similar to this already: https://www.ascensiadiabetes.com/eversense/ https://www.ascensiadiabetes.com/eversense/ I'd get a tattoo, but I'm not interested in anything that has to be removed
  - yazaddaruvala 4y ago
    My understanding is this isn't possible because of the size/invasiveness. We need significantly better power sources that are either have lots of energy capacity, xor some energy capacity and able to charge while inside the body. Either version they need to be small enough that the procedure is cheap and the body doesn't reject the device. Additionally, anything like this would need to go through regulatory bodies for medical devices, and would not be approved for "healthy people" (i.e. significantly reduced adoption). Roughly putting us back to non-invasive.
- refurb 4y ago
  But then youre still dealing with measuring it penetrating the skin and all the variables that have to with it - sensor position, changes in temperature and blood flow, etc. We do it with pulse ox pretty well, but +\- 2-3% O2 sat and repositioning isnt a big deal. Having the watch move and misreadibg blood sugar can have serious consequences.
- refurb 4y ago
  Its a valid ideal but then you get into the fun world of toxicity, pharmacodynamics, dosing, inter-patient variability, compliance, drug-drug interactions. A can of worms.
mortenjorck 4y ago
The article mentions pre-diabetes screening as a potential application - could that plausibly be accomplished with less precision than is required for a bolus calculation?
TylerE 4y ago
I mean, it's the no stick part that's tricky. Small, watchface size glucose monitors that transmit wirelessly exist. They're just...disposable. Wear it for about 14 days... stick on another $150 device.
- wpietri 4y ago
  There is also the "invasive" part. If you're talking about CGM devices like Freestyle and Dexcom, they include a probe that pokes through your skin.
  - efficientsticks 4y ago
    Type 1 diabetic here, they’re not close to as invasive as insulin pumps and a lot of people use those. Honestly the real problem with CGMs is they aren’t accurate enough, and they bias high - I suspect so they can pretend they eliminate hypos more than they do. (Having said that, CGMs are well worth it even with those drawbacks. They’re only invasive on the initial application.) But I do applaud the team for working on the technology nonetheless.
    - fwungy 4y ago
      Any info on how much they're biased?
      - kamens 4y ago
        I'm not so confident on this bias claim. Biasing high would be an extremely risky decision for a glucose monitoring company, because it encourages insulin dosing which could be deadly. Biasing low would be the move if scared of risky downsides.
      - efficientsticks 4y ago
        When performing calibrations I typically found my (accurate) blood finger prick monitor to be 1 mmol/L lower. It’s not enough of a difference to be deadly, as the sibling comment suggests, and the clinician guidance is to always eat something if you are feeling hypo, and at or near that level. The upshot is that I set my CGM-linked pump to target 1 mmol/L higher to compensate.
kerpotgh 4y ago
[dead]

copperx 4y ago

I wonder whether there are any shortcuts like ingesting a substance to make the glucose visible by the sensor.

Cshelton 4y ago
Bingo. Noninvasive, with any sort of accuracy that can be used for anything, is most likely impossible. So back to "invasive"... what can we do... embedding a device under your skin, right around where your watch goes, that can be powered and communicate with the watch. While also making sure the device is not rejected by the body and stays in place.. AND has no other adverse side-effects. I think this is where it'll ultimately end up for the people who want it. sub 10 min procedure to put it in and should last for quite some time (maybe dissolves after 5 years for upgrades?)
- blendo 4y ago
  Add heart rate, blood pressure, and blood oxygen level. Have the Apple Watch call 911 if things go pear-shaped. Profit.
  - Nekhrimah 4y ago
    > Profit. Well yes, but also substantial life saving interventions become possible.
  - Spooky23 4y ago
    They should profit. That will save millions of lives.
- MengerSponge 4y ago
  Eversense sells something similar to this already: https://www.ascensiadiabetes.com/eversense/ https://www.ascensiadiabetes.com/eversense/ I'd get a tattoo, but I'm not interested in anything that has to be removed
- yazaddaruvala 4y ago
  My understanding is this isn't possible because of the size/invasiveness. We need significantly better power sources that are either have lots of energy capacity, xor some energy capacity and able to charge while inside the body. Either version they need to be small enough that the procedure is cheap and the body doesn't reject the device. Additionally, anything like this would need to go through regulatory bodies for medical devices, and would not be approved for "healthy people" (i.e. significantly reduced adoption). Roughly putting us back to non-invasive.
refurb 4y ago
But then youre still dealing with measuring it penetrating the skin and all the variables that have to with it - sensor position, changes in temperature and blood flow, etc. We do it with pulse ox pretty well, but +\- 2-3% O2 sat and repositioning isnt a big deal. Having the watch move and misreadibg blood sugar can have serious consequences.
refurb 4y ago
Its a valid ideal but then you get into the fun world of toxicity, pharmacodynamics, dosing, inter-patient variability, compliance, drug-drug interactions. A can of worms.

mortenjorck 4y ago

The article mentions pre-diabetes screening as a potential application - could that plausibly be accomplished with less precision than is required for a bolus calculation?

TylerE 4y ago

I mean, it's the no stick part that's tricky. Small, watchface size glucose monitors that transmit wirelessly exist. They're just...disposable. Wear it for about 14 days... stick on another $150 device.

wpietri 4y ago
There is also the "invasive" part. If you're talking about CGM devices like Freestyle and Dexcom, they include a probe that pokes through your skin.
- efficientsticks 4y ago
  Type 1 diabetic here, they’re not close to as invasive as insulin pumps and a lot of people use those. Honestly the real problem with CGMs is they aren’t accurate enough, and they bias high - I suspect so they can pretend they eliminate hypos more than they do. (Having said that, CGMs are well worth it even with those drawbacks. They’re only invasive on the initial application.) But I do applaud the team for working on the technology nonetheless.
  - fwungy 4y ago
    Any info on how much they're biased?
    - kamens 4y ago
      I'm not so confident on this bias claim. Biasing high would be an extremely risky decision for a glucose monitoring company, because it encourages insulin dosing which could be deadly. Biasing low would be the move if scared of risky downsides.
    - efficientsticks 4y ago
      When performing calibrations I typically found my (accurate) blood finger prick monitor to be 1 mmol/L lower. It’s not enough of a difference to be deadly, as the sibling comment suggests, and the clinician guidance is to always eat something if you are feeling hypo, and at or near that level. The upshot is that I set my CGM-linked pump to target 1 mmol/L higher to compensate.

kerpotgh 4y ago

[dead]