Ask HN: Why no one has built Body Fat Tester Machine Startup?
The existing methods seem unreliable and not consistent. Why no startup has worked in this field?
Perhaps an opportunity to sell spare MRI time paired with ML for providing this product?
https://en.wikipedia.org/wiki/Dual-energy_X-ray_absorptiomet...
Is there any particular reason why at home device cannot be produced which is highly consistent and accurate?
I’ve done DEXA and the measurements between the two were very close
Smartwatches like the Samsung Galaxy Watch 5/6, which use wrist based sensors to measure body fat, are convenient but not very reliable. They typically have an error margin of ±3–5% (e.g., reading 17.5%–23.0% for someone who is 20%) because they only measure your torso and are sensitive to your hydration levels. A 2022 study in Obesity Facts by Kim et al. found a mean error of 3.8% for Samsung's BIA compared to a DEXA scan. The U.S. Navy Method, which uses a tape measure for your neck, waist, and hips, is a cheap alternative but has a ±4–6% error margin (e.g., 16.0%–24.0%) due to inconsistencies in measuring and variations in body shape. A 2019 study in Military Medicine by Hinton et al. reported a mean error of 4.2% for this method in men.