4 ms·

That's because it's bullshit. To clarify, the actual science they did is interesting (to me at least, as someone not in the field). The paper is linked here: h

• sheepscreek 2y ago
  > they didn't do any testing, at all, with actual humans (or animals) They did mention in the article that clinical trials are on-going.
  • hn_throwaway_99 2y ago
    Correct, but I'm extremely skeptical, and that sentence had my bullshit detector alarms screaming even louder: > “We have a minimum viable product that’s already being used in clinical trials, and while there’s more work to be done, we’re much closer to a full marketable device,” Shaker said. Absolutely no information about what this "clinical trial" entails, or what phase it was. Most importantly, to get an initial assessment of the accuracy of the device, no clinical trials are necessary - you simply need to do a test that compares the blood sugar reading from the device against the current gold standard, most likely first in some animal model. If their device was really as far along as the title and quotes are implying, they would be showered with so much money it would make the Theranos peak valuation look small. The only evidence they've provided (which, again, I'm not saying is insignificant) is that the "metasurface" they have developed enhances the resolution and sensitivity of a radar system against a beaker of water.
    • azinman2 2y ago
      But they have a minimal viable product! It’s viable! /s
      • caseyy 2y ago
        Well… yes. Looking at the article, it satisfies the MVP requirements for getting grant funding. A bar GlucoWatch cleared more than two decades ago, though it could never quite clear the bar of clinical viability. Theranos also had an MVP in this sense :)
        • cb321 2y ago
          I have probably seen "viable" overinterpreted 100s of times by now. Perhaps we need to re-interpret the 'V' in "MVP" as "VC-investable" or perhaps replace it with 'I' to be I)nvestible which gets you a more pronounceable "MIP" (and maybe, just maybe highlights uncertainty since all investment carries risk)? Happens to also abbreviate "Multum In Parvo" (Latin for "many/much in little") which is not even that far off from the semantic. ;-)
          • Terr_ 2y ago
            Indeed, and for different audiences, you need to be ready to present your MIP with different levels of detail. I propose we call this new process... MIP mapping. :p https://en.wikipedia.org/wiki/Mipmap https://en.wikipedia.org/wiki/Mipmap
            • caseyy 2y ago
              Indeed, producing (or “rendering”) a low quality MIP is easiest. :p
    • sheepscreek 2y ago
      It’s fair to be skeptical. Personally, I’ll believe it when I see it in action. There’s likely an unexplained catch or they wouldn’t have shared any hard data in the paper. One possibility is that they want to sell this technology to a big company without publicly disclosing all their trade secrets. However, this research could have been sponsored by a public grant, which would have compelled them to share some information. Therefore, they published a paper that appears more like a patent application than a research paper with solid data. It’s still noteworthy that it was published in Nature.
• m463 2y ago
  I thought apple was trying to get glucose monitoring approved https://en.wikipedia.org/wiki/Noninvasive_glucose_monitor#Near-Infrared_spectroscopy https://en.wikipedia.org/wiki/Noninvasive_glucose_monitor#Ne... (different technique)
• raphman 2y ago
  Thanks for calling this out. That was also my impression after having skimmed their paper: the only link to glucose monitoring is that the authors mention a few papers on the topic to motivate their research. And looking at the papers they cite, I see little evidence that this approach could work in practice in the near future. Most of the citations [2, 15, 16] are to their own work, which did not look at glucose monitoring in the human body. This is not my field of expertise, and maybe I am misunderstanding the papers. But it seems that there is little evidence that non-invasive glucose monitoring via measuring dielectric properties works reliably in practice. No in-the-wild studies, no investigation of potentially confounding factors. Take for example citation 22 from the paper. A study where the authors propose a new antenna design. They seem to measure how the pancreas changes size during insulin production by monitoring its dielectric properties. IIUC, they look for a dip in the frequency spectrum caused by absorption of a certain frequency band. But their measurements show an even larger effect when measuring on the thumb instead of the pancreas. This effect is not explained at all. (My guess: after having patients fast for 8-10 hours, giving them glucose will have an effect on the whole metabolism, resulting in higher blood flow, and that's what they measured). Also, while they operate the antenna in the GHz range, they use a cheap USB soundcard (sampling rate 44.1 kHz) for capturing the signal. I did not understand this at all. They also repeatedly use the term "dielectric radiation". Seems to be a rather uncommon term? The "machine learning algorithms" mentioned in the title seem to be a simple linear regression? They claim an accuracy of ~90% and show some sample results. The complete study data is only available upon request, however. [22] S.J. Jebasingh Kirubakaran, M. Anto Bennet, N.R. Shanker, Non-Invasive antenna sensor based continuous glucose monitoring using pancreas dielectric radiation signal energy levels and machine learning algorithms, Biomedical Signal Processing and Control, Volume 85, 2023, 105072, https://doi.org/10.1016/j.bspc.2023.105072 https://doi.org/10.1016/j.bspc.2023.105072
  • bobmcnamara 2y ago
    I assume there's an RF mixer somewhere in there. Edit: read the paper, now more confused
    • raphman 2y ago