Sigh...

LIGO detects length changes of 10^-18 m, or attometers, not femtometers, which are a thousand times longer. (https://www.ligo.caltech.edu/page/facts) But this does not matter at all, because this is not resolution of the body image, but the size of the vibration on the speaker. That's a technical data point that I don't see any reason to include in this presentation other than to cause this exact confusion.

The video looks in general like it's trying to impress by giving a lot of incidental information about how the device works while being very light on what it would be able to actually see -- e.g., it doesn't matter how many gigabytes your device collects if the resulting image is blurry.

Compare the website of LIGO (https://www.ligo.caltech.edu/page/facts), which also has a lot about the technical marvels (huge vacuum tubes! precision engineering!) but crucially includes the goal of this all.

The authors report explicitly state that the system's resolution does not match clinical CT or MRI. The elevational resolution is 15–25mm, meaning each slice is effectively a thick 2D section rather than the fine isotropic 3D volume MRI provides. MRI also delivers far richer soft-tissue contrast; this device produces only three contrast types (reflectivity, speed of sound, attenuation), and because it uses a low 1MHz frequency, the reflection images come primarily from tissue boundaries rather than from internal tissue texture.

I could see this being valuable for adipose tissue mapping or fatty-liver monitoring at a large scale, as the machines would be significantly cheaper, but this isn't some revolutionary magic bullet like the Twitter post is insinuating.

For context I'm in engineering consultancy, so by no means an expert but I probably have enough experience to be on the other side of the Kruger curve dip.

Passive sonar in the naval sense means listening only, not emitting. Do you mean imaging that relies solely on acoustic energy already present and emitted by the body? If so, then generally no. You have two types of "passive" imaging. First would be hardware-passive, as in MR elastography (most common), where the patient wears a transducer pad, and vibration is actively generated by a driver. You've then got algorithmically-passive, which is more analogue to passive sonar, reconstructing tissue stiffness from ambient broadband vibration without the emitted probing pulse, but that is very much entirely academic. I guess the question would be, why is it worth pursuing when you have something like optical coherence elastography (OCE) for non-invasive profiling. Doing it using noisy ultrasound method becomes redundant. There are other methods, but the outcome is the same.

Generally (this is true for all systems, not just humans) you need to induce energy into it to more effectively measure it's output. Think of it like a bell - if I want to hear the note it produces, it's much easier to hear what this is if I ring it with a hammer. Granted, it will be "passively" resonating to a point where, with a sensitive enough sensor, I could probably pick up the output without the hammer - but that is a pointless problem to solve. I could hit a bell with a soft hammer a million times over without causing damage to it. The lifetime of the person hitting it with a hammer is far shorter than the accumulative damage to the bell before it breaks. The same is true for humans. You could effectively run a very low-energy, 60Hz vibration through a person (which is how the pads work) for multiple lifetimes before it would cause significant damage, so there comes little point in solving that problem. As such, true "passive" imaging is functionally pointless if your outcome is "safely image a patient". You're overengineering your solution to solve a problem that is only relevant if your patient was planning on living for 1000+ years.

private equity

LNT does also damage, as people refusing necessary CT scans or countries switching of nuclear power because of fear.

LNT is the null hypothesis. No one disagrees a linear model fits the data very well in high doses. If you want to argue that model doesn't work in low doses, you need a model with more parameters and sufficient data to fit it. The issue is that, at these low doses we want to differentiate, we're also looking at effect sizes that are hard to separate from noise, and sampling biases that are hard to erase. There's still lively and ongoing debate.

We also used to x-ray people for shoe fitting/for lulz.

https://en.wikipedia.org/wiki/Shoe-fitting_fluoroscope

We also put lead in paint and gas because why the fuck not? Asbestos pretty much everywhere because it was a miracle material, etc. for decades, and we're still paying the price

> But I do agree with your point, these days, I hope we're better about studying the potential dangers of current technologies we use.

Sorry, but this is just pure "Gell-Mann amnesia effect" vibe to me. I mean, you've just brought up a perfect example yourself! What kind of mental gymnastics does it take to still hope that this time it's not like that?

I don't wanna start the whole "vaccines cause autism" thing and whatnot, and surely you shouldn't avoid ultrasound just because of irrational fear of some yet undiscovered side-effects, but it's really amazing, how people tell fun stories about how common was the narrative about major war being very unlikely in "modern days" (because who would dare to do that with this kind of technology!) right before WW1, and then conclude with firmly believing that these days (after WW2) it sure won't happen, because humans are not that dumb. And my point is, that perhaps it indeed might have been a bit less likely, if people didn't believe that it is so unlikely to keep stepping on the same rake.

> The first video appears to be real

The video is clearly from Midjourney /s