Noninvasive imaging method can penetrate deeper into living tissue | Hacker News Reader

Noninvasive imaging method can penetrate deeper into living tissue

news.mit.edu

199 pointsrbanffy24 comments

Mary Lou Jepsen of openwater.cc [2] has managed to image neurons and other large cell structures deep inside living bodies by phase wave interferometry and descattering signals through human tissues, even through skull and bone[1] using CMOS imageing chips, lasers, ultrasonics and holographics. A thousand times cheaper fMRI. She aims to get to realtime million pixels moving picture resolutions of around a micron. Eventually she will be able to read and write neurons or vibrate and someday maybe burn cell tissues to destruction.

There are later presentations at the website where the technique is better described and visualized, but [1] is a good quick place to start and judge if you want to study their brainscanner in more depth. There are patents and a few scientific publications [3] that I'm aware of, but mostly many up to date talks with demonstrations by startup founder Mary Lou [4]. And recently she is open sourcing[5] parts of the hardware and software on github [6] so we can start building our own lab setups and improve the imageing software.

[1] https://www.youtube.com/watch?v=awADEuv5vWY

[2] https://www.openwater.health/

[3] https://scholar.google.com/citations?hl=en&user=5Ni7umEAAAAJ...

[4] https://www.youtube.com/watch?v=U_cHAH4T8Co

[5] https://www.youtube.com/watch?v=hNFQtpNHufk

[6] https://github.com/OpenwaterHealth

When the best you can come up with are TED talks and videos on a cryptocurrency youtube channel - and the only claim to fame she has is that she co-founded a miserable failure of a project (One Laptop Per Child) there are serious credibility issues.

Is there any peer reviewed research?

Is the collaborating with any established, published, respected-in-their-field researchers?

After a decade of research, have they shown any actual pathway or mechanism of treatment?

They're not helped by the FDA politely saying "you haven't actually said how you intend for this thing to be used" and "your protocols don't actually specify why they are safe" and "no, we're not going to give you shit, because you haven't even done animal studies or clinial studies" and "no, you can't just say that your device has no risk of injury or death": https://github.com/OpenwaterHealth/opw_regulatory/blob/9e151...

She hasn't even done any animal studies, and went to the FDA seeking approval for a device...

These people are a bunch of clowns. The FDA has to explain to them that they could do thermal testing by inserting a themocouple between the device and fake skin on a testing device.

From the OpenwaterHealth/opw_neuromod_sw README: https://github.com/OpenwaterHealth/opw_neuromod_sw :

> open-LIFU uses an array to precisely steer the ultrasound focus to the target location, while its wearable small size allows transmission through the forehead into a precise spot location in the brain even while the patient is moving.

Does it use one beam of the array to waveguide another beam? For imaging or for treatment?

From "A simple technique to overcome self-focusing, filamentation, supercontinuum generation, aberrations, depth dependence and waveguide interface roughness using fs laser processing" (2017) https://www.nature.com/articles/s41598-017-00589-8 https://scholar.google.com/scholar?start=10&hl=en&as_sdt=5,4... :

> We show that all these effects can be significantly reduced if not eliminated using two coherent, ultrafast laser-beams through a single lens - which we call the Dual-Beam technique. Simulations and experimental measurements at the focus are used to understand how the Dual-Beam technique can mitigate these problems. The high peak laser intensity is only formed at the aberration-free tightly localised focal spot, simultaneously, suppressing unwanted nonlinear side effects for any intensity or processing depth. Therefore, we believe this simple and innovative technique makes the fs laser capable of much more at even higher intensities than previously possible, allowing applications in multi-photon processing, bio-medical imaging, laser surgery of cells, tissue and in ophthalmology, along with laser writing of waveguides.

I remember she did a talk at the Long Now Foundation a while back. My take away - which could be completely wrong - was that they saw the potential to read and write memories.

She can't even answer the FDA on "explain how this will actually treat anything"

https://github.com/OpenwaterHealth/opw_regulatory/blob/9e151...

Interesting, but probably not cheap:

> This is achieved by having more than 0.5 megawatts peak power

and

> The input peak power up to 1.60 MW (350-nJ pulse energy)

https://www.science.org/doi/10.1126/sciadv.adp2438

I wonder how much it affects/fries the tissue. HHG also has issues with attenuation.

The paper abstract also says previous state of the art was 300um not 200um like in the press release.

The question is, how long is that power needed.

Edit: "Ultrashort pulses" so likely not much. Because indeed, you cannot burn the human. So the energy consumption likely won't be a problem. Except that it might make the machine expensive to build.

In fact, GP quotes "350 nJ" which is a tiny fraction of a watt-second.

Bulk ordered, that is probably a ~$50k laser. Not nothing, but on the scale of medical device costs, not terrible.

300 to 700 micrometers? So almost nothing to still almost nothing.

Could anyone tell a difference in the before and after images, other than one was grey and the other was blue? Structurally, they looked identical to me.

Edit: internet tells me human cells are around 25um, so I guess you can tens of cells deep

I think this is just one "slice" of a full scan. The one in blue has a lot more detail, and less noise.

Does it? I zoomed in on my phone as far as it would go, and could still barely tell the difference. Unfortunately the difference in hue makes it harder to tell if there's any actual difference, because it swamps everything else.

Since they claim a more-than-doubling of visibility depth over SoTA, I'm surprised they didn't pick a visual that demonstrates that quite radical improvement, over the apparently marginal improvement in resolution at depths which can already be managed.

I also didn't distinguish any difference, don't see how the choice of color improves the visualization. Honestly it comes across as a bit deceptive.

The blue one is pretty significantly less noisy, and the animation showing the two 3D reconstructions is much clearer. The noise reduction probably has a lot to do with that.

It allows you to study things in-situ that are just under the skin without opening things up. Useful!

Radiologists spend 12 years studying to be able to interpret imaging - as a lay person who has spent half that duration looking at scans most days, I still can’t really form any sort of accurate picture of what’s going on.

I have a suspicion that radiologists themselves may not know what's going on. Essentially if you give a scan (with cannot be easily interpreted by a semi-trained layman) to more than one radiologists, you may get several different interpretation.

Can some radiologist reading this confirm what I'm saying?

Radiologists probably like the gray noisy image better anyways!

Yeah this isn't replacing MRI anytime soon.

>excitation of NAD(P)H at 1100 nm.

Better to understand this in context of the general techniques like this:

http://en.wikipedia.org/wiki/Near-infrared_window_in_biologi...

What about NIR spectroscopy? It penetrates the bio window at >1200 microns now…not sure how this compares? I mean signal processing wise, I would assume NIR does the trick.

Sorry meant 1200nm*

A better title and subtitle is needed though... "Noninvasive" and "Using high-powered lasers" might sound a little bit an oxymoron...

Hmm that before/after photo looks like they spent about 5 minutes in Photoshop tweaking the curves. Is it really meant to demonstrate an improvement?

Reply on news.ycombinator.com