LESSWRONG

Sep 26, 2016

13 comments

Apple's iPhone 7 Plus decided to add another lense to be able to make better pictures. Meanwhile Walabot who started with wanting to build a breast cancer detection technology released a 600$ device that can look 10cm into walls. Thermal imaging also got cheaper.

I think it would be possible to build a 1500$ device that could combine those technologies and also add a laser that can shift color. A device like this could bring medicine forward a lot.
A lot of area's besides medicine could likely also profit from a relatively cheap 3D scanner that can look inside objects.

Developing it would require Musk-level capital investments but I think it would advance medicine a lot if a company would both provide the hardware and develop software to make the best job possible at body scanning.

2 points

Comments

sorted by

Highlighting new comments since Today at 9:18 AM

All comments loaded

James_Miller

3 points

3 points

3 points

I think that many physiotherapists could do a better job if they would have body scanners.

The BMI is a horrible metric and having cheap body scanners would move us past the BMI and provide us with better targets for weight management.

Given that, wouldn't having lots and lots of these scanners massively increase medical costs by creating many false positives?

In many cases I wouldn't need to go to the doctor if a good body scanner can tell me what's up with me. If the scanner can tell me whether my teeth are alright, I don't have to go to the dentist.

If I can get a body scan for mammogram from a person who isn't a breast surgery salesman as in the status quo, a false positive is also less likely to get me to do risky treatment.

3 points

Good_Burning_Plastic

0 points

The BMI is a horrible metric and having cheap body scanners would move us past the BMI

We have had cheap bathroom scales measuring body fat percentages (not terribly accurately, but still better than guessing from the BMI) for a while; if those didn't "move us past the BMI", why do you think a device two orders of magnitude more expensive would?

0 points

0 points

Cheap scales don't measure body fat uniformly. They ignore arm composition. For the purposes of standarization they give different answers than the expensive devices used in clinical studies.

Fitness studies also measure more than body fat. They measure the circumference of various body regions. I don't think a measurement that doesn't take into account the shape of a body produce a good answer.

why do you think a device two orders of magnitude more expensive would?

Most medical devices that set standards aren't very cheap. Very cheap devices give nobody an incentive to run the studies

0 points

0 points

Unfortunately, I don't think we have the regulatory regime to sell them to consumers. At least, not in the US and Europe, and not while making available any information on how the scanner might produce medically relevant data.

Another Musk-level capital investment might be needed to solve this hurdle.

0 points

0 points

Legally I think the scanner might not be allowed to tell you whether you have breast cancer but I think it might be allowed to show you a pretty 3D picture.

0 points

0 points

The division into a scanner, and a person who interprets its results, is arbitrary. Both are subcomponents of a single apparatus.

If the scanner produces a hard to interpret picture, and an expert human interprets it (or publishes instructions for doing so), then maybe the scanner itself would be judged legal - although I expect judges would apply a standard similar to "does it have significant noninfringing uses?"

If the scanner attaches to each image a probability of breast cancer, encrypted with a secret key, and the expert human is merely decrypting the result, then the scanner would probably be prohibited too.

These are two points on a smooth gradient where the scanner outsources more or less work to the human. Where along it does the scanner become illegal? Probably at the point someone decides to stop it to make a point.

0 points

0 points

The division into a scanner, and a person who interprets its results, is arbitrary. Both are subcomponents of a single apparatus.

23andMe can give you your DNA information but they can't give you probabilities for having an illness. On the same token I would expect there a way to give you a clear image of the body scan.

The Apple watch is allowed to tell me my heart rate but not that I have heart disease.

although I expect judges would apply a standard similar to "does it have significant noninfringing uses?"

Practically it isn't so much the decision of the judge but the decision of the FDA. As far as the FDA goes I think it might be okay with the software telling certain patients: "The software detect something that might be abnormal, please go to your doctor to get checked".

0 points

0 points

In the case of 23andMe, the part the FDA chose to ban was that of anyone telling you that a certain allele was correlated with cancer. The reason they did so, and did not ban 23andMe from telling you what alleles you have, wasn't that they don't have the legal power to do otherwise, but that they correctly believe very few people would be able to analyze alleles themselves, and so only the analysis part had to be banned.

In any new product, I expect they will (attempt to) ban whatever they have to to achieve the goal of people not receiving significant medical information about themselves from outside the official medical system, information that would be commonly used to make medical decisions.

Which is exactly why they might be OK with any product that told people "you might have a problem, go to a doctor" even if the true thing detected by the product was "you have a problem, period". If a product turns out to cause many people to believe they have a problem without going to a doctor, the FDA would attempt to ban it.

0 points

0 points

Which is exactly why they might be OK with any product that told people "you might have a problem, go to a doctor" even if the true thing detected by the product was "you have a problem, period". If a product turns out to cause many people to believe they have a problem without going to a doctor, the FDA would attempt to ban it.

Yes, I think there a way to manage the situation that the FDA approves the product.

In the long run it's possible to actually run the studies that are required to diagnose certain illnesses.

0 points

0 points

These studies cost tens or hundreds of millions of dollars, take months or years to run, and then the process of getting the FDA to approve your product takes more millions of dollars and more years. At the end you have approval for a particular product for a particular purpose, and if you want to launch v2.0 or to use the product's measurements in a new kind of analysis for a new disease, you need to go through the approval process again.

This is entirely incompatible with quickly evolving technology in a market with many small / new players trying out various ideas. You end up with only a company the size of Apple being able to launch an Apple Watch, and even then it only being able to tell you your heartrate.

If the Apple Watch had advanced sensors for e.g. blood spectroscopy, and people could write programs that would tell you your blood sugar rates, and other people would write apps to tell if you had diabetes or advise you on insulin dosage, and people could side-load these apps without going through the Apple Store (maybe Android would be more appropriate), do you think the FDA wouldn't step in?

0 points

0 points

The FDA might shut down the independent App developers but I don't think the would shut down the Apple Watch for being able to run those Apps.

These studies cost tens or hundreds of millions of dollars, take months or years to run, and then the process of getting the FDA to approve your product takes more millions of dollars and more years.

I don't think a breast cancer screening study would cost hundreds of millions. You could go to a hospital that does breast cancer screening and let every patient scan themselves with a device while they are in the dressing room.

Then you can gather the statistics of whether the software can reliably produce the same diagnosis as the existing technology.

This is entirely incompatible with quickly evolving technology in a market with many small / new players trying out various ideas.

Yes. Especially at the beginning you would have most of the software being developed by the company who sells the device. Business wise that means it might be very worthwhile for a company who develops this technology because they might have a monopoly on the marketplace.

0 points

0 points

I don't think a breast cancer screening study would cost hundreds of millions.

That's the wrong analogy; we're talking separately selling a device that tells you some metric, not a diagnosis like breast cancer. So presumably you have to compare it against some existing way to measure that metric.

I don't know how much testing & FDA approval normally costs for nontrivial "metric only" devices (e.g. glucometers), but on the outside view, I doubt it's cheap as you say, because everyone in the system has incentives to raise the price, just like for drug trials.

Especially at the beginning you would have most of the software being developed by the company who sells the device. Business wise that means it might be very worthwhile for a company who develops this technology because they might have a monopoly on the marketplace.

And product wise that would result in the same kinds of products we have now, i.e. extremely conservative and slow moving. And approved and marketed for just a few uses, instead of being open-ended like you suggested.

0 points