20
votes
Spotify co-founder Daniel Ek hopes his latest brainchild, the Neko Body Scan, will revolutionise healthcare
Link information
This data is scraped automatically and may be incorrect.
- Title
- I tried the £299 full-body scan that checks health risks in minutes
- Authors
- Linda Geddes
- Published
- Sep 20 2024
- Word count
- 937 words
So, it's a pan-scan. A tech-bro startup pan-scan.
This is a bad idea. And the reason for it is that human bodies are surprisingly non-uniform; pan-scan everybody, and you'll find something "wrong" in nine out of ten people that you could, theoretically, "treat". But in almost all of those people, those "findings" will never materialize in a symptomatic disease, and treatment would be all downside—side-effects only, no benefit. (To say nothing of the effects on mental health of knowing there's something "wrong".) This is why doctors don't already do this. Yeah, we could run everyone through a full-body CT and battery of blood tests—there's probably not capacity for the entire population, but it's not in principle difficult—but the net effect on population health of doing that is actually negative.
I’ve read about that before and understand that it’s often a problem, but I still wonder if there’s some way to use the additional information for good. It seems like it ought to somehow be possible to learn how to do something like watchful waiting?
I also wonder if a “before” picture could be used to avoid misdiagnosis: “this anomaly looks strange, but it’s not actually related; it’s always been that way.”
Preemptive full diagnostic scans are generally a bad idea for Bayesian reasons (XKCD for reference). Basically, if a test has a 3% false positive rate but the disease only occurs in 1% of the population, 66% of the test results will be false positives if you scan the whole population!
This is usually less of a problem in standard practice because smart doctors should only be ordering tests for suspected diseases. The exceptions are things like allergy tests which have stupidly high false positive rates but that's a whole different can of worms...
I wonder if this would be the kind of task where ai could be of use. If everyone we're having these scans (or merely a vastly higher number of people) more frequently, ai might be able to identify patterns and flag certain cases for doctors to take a closer preemptive look at.
AI or any other method makes zero difference to the fundamental statistics. As long as the false positive rate is higher than the disease's incidence rate amongst the general population, it's not a great idea to prescreen everyone.
In some ways, this reminds me of Theranos. My degree is in biomedical engineering, so I feel very confident saying there are certain things we just can't or shouldn't do. I'm sure their tech works, but it's a bad idea to always use it on everyone unless it's really cheap and easy to repeat (false positives are much less of a problem if you can test three+ times).
I get the theory of the problem you describe, since it introduces all kinds of new concerns: feeding hypochondriacs with worry, risking affluent patients getting care for minor ailments, and a considerable number of false positives leading to wasted resources.
However, I do think that needs to be weighed against the other side. We might be able to flag serious diseases in earlier stages, or early signs of cancer before it can metastasize. We could establish personal baselines for individuals, and later show any deteriorations over time. And with all that extra data, we could presumably improve our statistical modeling to reduce the false positives (perhaps not overall, but on an individual level).
The initial issues could be addressed or offset through some targeted policy decisions. For example, mandating that we still prioritize treatment based on assessed risk, so that nobody can get ahead in line regardless of how rich they are. I say that as somebody living in a country with socialized healthcare though, so that's already the expected norm here. Not everything should necessarily be shared with the patient either, if they're likely to misinterpret or try to self-treat the problem.
The biggest issue seems to be that of resources. There aren't enough machines, personnel, time, or money to scan everybody. Presently, that makes it a good decision to not spend those limited resources in areas where they're unlikely to do much good. However, as technology improves, the equation may shift. New techniques often prove cheaper and more portable. Future scanners may not require dedicated technicians. Improvements to accuracy, or advances in data modelling may also reduce false positives. So at some point, it very well could be feasible offer routine scans to everybody. If and when that happens, we'll need to weigh the tradeoffs of performing such scans.
It seems to me that more data is generally a good thing, both in diagnosing and staving off problems. While there are always complex human factors to consider, I don't believe any of them are unsolvable. As with anything else in healthcare, it requires thoughtful consideration, empathy for the individual, and smart decision-making to find the right balance between the challenges and the potential benefits.
If the tests are easy and cheap to replicate three+ times, the false positive rate is much less of a concern. With a one-off scan, there are statistical limits though, and no amount of algorithm will fix the underlying problem caused by false positives. As you mention, there are always resource constraints, so we have to weigh the benefits against the costs.
I’ve often wondered about getting a full body MRI scan every 5 years or so to catch early signs of issues and potentially cancer, is that reasonable?
Not really. If you're in good health, the test is more likely to give you a false positive than help a doctor find anything wrong. This depends a lot on your age and health though. If you're looking for some signs of a congenital condition that runs in the family, that's different.
I think so. The idea that a regular checkup is bad because potential false positives ignores the chance of genuinely catching something serious at an early stage. Stage 1 cancer is much easier to fix than stage 4. Some people just enjoy poking holes at anything new and unfamiliar, especially if it comes from someone, or a category of people (see OP 'tech bro'), that they dislike.
"Neko" means "cat", right? So it's a cat scan. Funny. Anyway, I hope they are successful. Automation will be key for health care in the West in the future.
I find this kind of funny, having just come from reading through this comment. https://tildes.net/~life/1iww/japan_faces_labor_shortages_and_demographic_crisis_as_elderly_population_hits_record_high#comment-do4b
I don't get it. The "CAT" in CAT scan stands for "computer assisted tomography". A tomography needs the subject to be cut in slices. The Neko scanner is not doing slices, its mostly a surface scan. So what's the joke?
Much like SCUBA or LAZER, people mostly don't know/care what the thing stands for, they just use it as a name.
"Neko" being Japanese for "cat" is the whole 'joke'. It's not a great joke.
Considering Spotify isn't in the medical field, I'm willing to bet this is NOT in fact Daniel Eks brainchild, and that he is really trying to pull a Musk and claim it is.
He’s in the headline on Tildes, but that was the subhead in the actual article. The article does mention the co-founder too:
And here’s how the company explains it on their website:
The media often has a tendency to focus on one leader at a company and in this case, the way it was posted on Tildes inadvertently exaggerated that effect. There’s a tendency for news articles to exaggerate the influence of people you already heard of because that gets more attention, and often, we fall for it.
And to some extent, that’s the job: famous founders will try to get attention for their company using their own fame.
Agreed, except for the entire US pharmaceutical industry, who is happy to sell you medication for your preventable chronic medical issue as long as your profit-based insurance company will fork over the hugely inflated cost.
I really hope this becomes a positive thing, they seem to have really good intentions. I fully expect that the US healthcare system would warp it into something grotesque to prioritize stock prices and de-prioritize people's actual well being.
I look at the philosophy of the us healthcare system as a superhero. It's designed to swoop in and save the day, not to avoid the problem in the first place.
A perfect example of this is the unhoused. These folks often don't have access to regular healthcare for things like psych care and insulin, but they can go to the emergency room at no cost to them and walk out having received temporary respite from the weather, patching up if they need it, and a month's worth of medicine.
I think more could be done on prevention, but there are some inherent limits that make providing healthcare largely reactive in any healthcare system. Even for something very cheap and effective like vaccinations, there are limits because people have to agree to get vaccinated. If they don’t come in to do it, nobody is going to make them. (We saw what happened when they tried to put more pressure during the pandemic.)
Similarly, you can’t make people take their meds or exercise or eat healthy or show up for a doctor’s appointment. The system is mostly about offering services to people who want them, and the people are often reluctant. They might not show up until things get so bad that they need the emergency room.
Though, it would certainly help if medical services were easier to get when people do want them.
I think you raise a good counter-point to the very thesis of the founder's statement: People may "want" a preventative one, but they in no way act as such. If American's truly wanted a preventative, proactive healthcare system, there would be a heck of a lot more exercising and given body mass trends in the US, it's hard to argue that we're trying to prevent any of the litany of diseases/disorders/issues that arise from a sedentary lifestyle.
I'm wary of making universal statements like that about people. Exercising isn't rare! Many people care about their diets, at least sometimes. People vary a lot, they can change their minds, and they can be inconsistent.
I appreciate the more even hand. Perhaps it's my proximity to the medical field where I feel more cynical about the proactiveness of the average consumer.
This dude's ruined enough already