Raspberry Pi thinks even $20 or $35 is way too much to pay for a computer. That's why they just announced the Raspberry Pi Zero, which they're happy to sell for a startling $5. That's right, $5. Not $5 on some sort of monthly installment plan or as a cloud-based monthly subscription service, but a straight up purchase of a programmable computer for only $5.
I wish health care had a Raspberry Pi to drive down its costs in a similar fashion.
For those of you who had not been familiar with Raspberry Pi, their mission is to get inexpensive computers in the hands of more people, especially children. They recognized that many families can't afford several hundreds or thousands of dollars for a computer, as you'd normally pay at Best Buy or Amazon. And they believe it is imperative to let more people -- again, especially children -- have the opportunity to experiment with programming.
Their first computer, the Raspberry Pi 1 Model B, was introduced in 2012 for $35, and was followed over the next few years with other models that were as low as $20. They made them cheaper and/or more powerful, but never more expensive. Not content with selling some 6 million of these low cost computers, and following the advice of Alphabet Inc's Chairman Eric Schmidt that it was "hard to compete with cheap," they developed and rolled out the Zero, figuring $5 was as low as they could go. No kidding.
The initial batch of the Zeros sold out within a day.
Sure, for the $5 you're not getting an Apple MacBook or a Microsoft Surface Book, but it is powerful enough to play games, connect to the Internet, and use its simple programming language to connect it to home devices, create your own games, maybe even build a robot or two. I'll probably be sticking to my PC -- I'm not likely to be building any robots anyway, much as they fascinate me -- but I'll bet there are a bunch of teens or even preteens who could do some pretty cool things with a Zero.
For $5, why not? It might be a great stocking stuffer or Hanukkah gift.
Meanwhile, in health care we're not surprised to read headlines like "Cost of Skin Drugs Rising Rapidly," reporting on a study that found prices of 19 brand name dermatologic drugs have risen 5 fold, on average, over the past six years, with most of those increases happening more recently. It'd almost be funny if we weren't the ones paying these huge increases, which of course we are, either directly or through higher health premiums.
I'm not going to pile-on any further about prescription drug increases, especially not after reading the analysis in Health Affairs by Kenneth Thorpe and Jason Hockenberry that suggests prescription drugs may not be quite the culprits we're getting used to thinking they are. No, there are plenty of parties in the health care system that make the health care equivalent of a $5 computer almost impossible to imagine.
We have, after all, a health care system in which heart patients apparently do better when the top cardiologists aren't around, because fewer things are done to them, and in which "inaccurate and unreliable tests" are resulting in unnecessary care, raising costs, and putting patients at risk, according to the FDA.
Sure, we have organizations like Diagnostics For All, which works on developing "low-cost, easy-to-use, point-of-care diagnostics," with particular interest in their use in developing countries. (A couple of months ago I might have cited lab pioneer Theranos as another example, but that's probably not such a good idea right now). Many people think mHealth is a particularly potent way to introduce low-cost care options, again mostly for developing nations. UNICEF is sponsoring The Wearables for Good campaign to help spearhead this kind of effort.
But let's face it: when it comes to health care, we are a developing nation; we just pay more for it than anyone else.
Even after ACA, we still have some 32 million non-elderly uninsured, some 44 million insured people avoid getting care due to concern about costs, only half of Americans rate the quality of our care as at least good, and by more objective measures our health care system is closer to the bottom than to even the middle when compared to other "developed" countries.
So where is, say, our $5 EHR? We can't even get EHRs that physicians like. Where is our $5 MRI? Their range in prices is eye-opening, but it's safe to say even the cheapest would make a full price computer look cheap. One could argue that we have $5 drugs, what with close to 80% of prescriptions filed being generic, but I don't see many people jumping up and down with excitement about how cheap their prescriptions are.
For that matter, where is our Mozilla, our Wikipedia, or our Linux, offering widely used free services of tremendous value?
In our health care system, we think we're winning if we slow the rate of growth, although it remains above overall inflation. Even new entrants seem to be more interested in getting their piece of the $3 trillion pie than in doing what Raspberry Pi is doing for computers.
Who is going to be our Eben Upton, our Jimmy Wales, our Linus Torvalds? Wouldn't it be cool if it ended up being one of those kids tinkering on his Zero?
Friday, November 27, 2015
Wednesday, November 18, 2015
Breaking Barriers
If you are literally starving to death, you can't expect a restaurant to feed you. If you are homeless and the forecast calls for sub-zero temperatures, you can't expect a hotel to put you up. But if you think you are having an emergency health problem, by law you can go to a (Medicare-participating) hospital emergency room to get evaluated and at least stabilized, even if you can't pay.
Similarly, I've heard many calls for a single payor health care system, but I've never, not once, heard anyone advocate that the government should pay for all our food or our housing.
For some reason, we think about health care differently, even from other life-sustaining needs like food and shelter.
I like reading predictions about how healthcare is soon to be radically disrupted as much as anyone -- Oliver Wyman's The Patient to Consumer Revolution and David Chase's new piece in Forbes are two of the latest -- but I worry that many of these ideas are like putting nicer lipstick on the pig.
Let me get to each of these in turn.
Having physician-run state medicine boards oversee who can practice medicine is usually positioned as a benefit to patients, supposedly ensuring that we get care only from qualified professionals. Some (all right, I) have argued that, whatever their intent, the temptation for such self-policing bodies to end up enabling a cartel is hard to resist.
Here's two examples of the problem:
State-based licensing is an artifact of an earlier, more geographically restricted day. The medical boards trumpet the new telemedicine compact to show they they are making progress about crossing state borders, but it falls woefully short of what technology allows. We're starting to have more options, and we'll increasingly want even more of them.
Similarly, I've heard many calls for a single payor health care system, but I've never, not once, heard anyone advocate that the government should pay for all our food or our housing.
For some reason, we think about health care differently, even from other life-sustaining needs like food and shelter.
I like reading predictions about how healthcare is soon to be radically disrupted as much as anyone -- Oliver Wyman's The Patient to Consumer Revolution and David Chase's new piece in Forbes are two of the latest -- but I worry that many of these ideas are like putting nicer lipstick on the pig.
Yes, innovations like digital health, Big
Data, and mHealth hold great promise, but even they eventually run into what most people might view as two
of the foundations of our health care system but which I fear may actually be
hitherto impenetrable barriers to change: "the practice of
medicine" and "the business of insurance." My calling them
barriers should come as no surprise to readers of my previous posts.
Let me get to each of these in turn.
Having physician-run state medicine boards oversee who can practice medicine is usually positioned as a benefit to patients, supposedly ensuring that we get care only from qualified professionals. Some (all right, I) have argued that, whatever their intent, the temptation for such self-policing bodies to end up enabling a cartel is hard to resist.
Here's two examples of the problem:
- Let's say I have a rare condition, and it turns out that the best physician to help me with it is in California. Or Germany, or India. Fortunately, modern technology allows me to consult with them via video, and increasingly would allow them to perform tests and even procedures on me from where they are. Unfortunately, under our current approach, unless they are licensed in the state I'm in, they can't help me.
- As health care accumulates Big Data and the AI programs to sift through it, someday soon such a program will figure out someone's obscure diagnosis and perhaps propose a novel treatment, one that isn't (yet) supported by clinical research or even medical theory. We may not be able to understand how the program concluded it was the right diagnosis and treatment, but that doesn't mean it won't be. But, of course, such a program can't practice medicine.
As I've written before, the more we learn about the body and how to keep it healthy, the more it may be that physicians may not be the best people to treat us in all cases. I'm not talking about physician substitutes like nurse practitioners but, say, geneticists, robotics experts, or microbiologists. If we subject all the coming advances to our existing ideas about who can "practice medicine," we will be missing out.
And no one that I know of is seriously thinking about how to "license" the inevitable AI-based experts. That human-centric point-of-view is natural, but will not survive the 21st century.
To be sure, proof of competence for our health care providers is to be desired, but state licensing is not the only, or the best, way to accomplish this. Board certification exams, for example, don't vary by state. Whatever proof we demand should be more universally comparable, more empirically/performance based, and more transparent. There's no reason that resulting "license" has to be geographically limited, or even limited to our traditional types of providers.
As for the "business of insurance," here are quotes from two dissatisfied customers in a recent The New York Times article about high deductible plans:
Instead of primarily protecting us from the risk of catastrophic expenses, we now expect health plans to pay for our routine care as well, tell us which providers we can see, negotiate discounts on our behalf, and help us manage our care. Maybe those are good things, maybe not, but what they are not are things that only health insurance could do. In fact, even helping us finance care is not something that only health insurance can do.
We could be thinking of different approaches to financing our health care. Why shouldn't anyone be allowed an HSA? Crowdsourcing and microloans are all the rage in financial circles, especially with the FinTech revolution to enable them. There's no reason these couldn't be used for health care expenses.
Or think of approaches analogous to life insurance, with payouts based not on death but on catastrophic health events -- a lengthy hospital or nursing home stay, for example. I'm pretty sure actuaries could price these, but I'm not as sure that many people would be willing to buy them.
Yet.
But none of that is going to happen if financing innovators have to worry about being considered in the business of health insurance, which would then impose a raft of requirements on them that would force them to look and act much like the health insurance we have now. And that'd be a shame.
Innovating in health care shouldn't just be about doing what we are doing with better technology, but must also be about rethinking what can help us achieve better health, who is truly best qualified to assist us with that, and what our range of options is to finance our health needs. We can't be limited by traditional notions about the practice of medicine or the business of insurance.
If we want to think about health care differently, let's really be different.
And no one that I know of is seriously thinking about how to "license" the inevitable AI-based experts. That human-centric point-of-view is natural, but will not survive the 21st century.
To be sure, proof of competence for our health care providers is to be desired, but state licensing is not the only, or the best, way to accomplish this. Board certification exams, for example, don't vary by state. Whatever proof we demand should be more universally comparable, more empirically/performance based, and more transparent. There's no reason that resulting "license" has to be geographically limited, or even limited to our traditional types of providers.
As for the "business of insurance," here are quotes from two dissatisfied customers in a recent The New York Times article about high deductible plans:
"Our deductible is so high, we practically pay for all of our medical expenses out of pocket. So our policy is really there for emergencies only, and basic wellness appointments."
"I will never be able to go over the deductible unless something catastrophic happened to me. I’m better off not purchasing that insurance and saving the money in case something bad happens.”Just what, exactly, do they think insurance is?
Instead of primarily protecting us from the risk of catastrophic expenses, we now expect health plans to pay for our routine care as well, tell us which providers we can see, negotiate discounts on our behalf, and help us manage our care. Maybe those are good things, maybe not, but what they are not are things that only health insurance could do. In fact, even helping us finance care is not something that only health insurance can do.
We could be thinking of different approaches to financing our health care. Why shouldn't anyone be allowed an HSA? Crowdsourcing and microloans are all the rage in financial circles, especially with the FinTech revolution to enable them. There's no reason these couldn't be used for health care expenses.
Or think of approaches analogous to life insurance, with payouts based not on death but on catastrophic health events -- a lengthy hospital or nursing home stay, for example. I'm pretty sure actuaries could price these, but I'm not as sure that many people would be willing to buy them.
Yet.
But none of that is going to happen if financing innovators have to worry about being considered in the business of health insurance, which would then impose a raft of requirements on them that would force them to look and act much like the health insurance we have now. And that'd be a shame.
Innovating in health care shouldn't just be about doing what we are doing with better technology, but must also be about rethinking what can help us achieve better health, who is truly best qualified to assist us with that, and what our range of options is to finance our health needs. We can't be limited by traditional notions about the practice of medicine or the business of insurance.
If we want to think about health care differently, let's really be different.
Wednesday, November 11, 2015
Someone Must Be On Drugs
As is probably true for many of you, I'm busy looking at health plan open enrollment options for 2016. I have to confess that for the past few years I've been guilty of just sticking with the same plan, so it has been too long since I've had to shop. Plus, I'm helping my mother pick her Medicare options for next year. All in all, I'm awash with health plan options.
I'm torn between thinking that the people designing the plans are extremely clever, have a perverse sense of humor, or were under the influence of psychedelic drugs at the time.
It's not that there aren't plenty of options. I've got different levels of HMO, POS, and PPO options, from multiple carriers. My mother has many choices of Medicare Supplements, with Part D options, as well as Medicare Advantage options (both HMO and PPO), each from multiple health insurers.
Nor is it that having choices is bad. Researchers have discussed the "tyranny of choice" (or "paradox of choice") for some time, meaning too many choices can be paralyzing for consumers. I have to admit that when I go to the cereal aisle in the grocery store I feel overwhelmed, but, whether it is cereal or health plans, I'd still rather than more choices than fewer choices.
It's just that, well, the options are so damn confusing. I was in the health plan business for a long time, and helped develop some of the first plan selection tools for consumers, But when it comes to evaluating some of the options now available, I find it practically impossible.
Austin Frakt recently wrote in The New York Times about this problem. He cited a few studies specifically on point about health insurance, such as:
I'm most frustrated with prescription drug coverage. Not that long ago, the only variables were the copays for generic versus brand drugs. Now there are often five or six different tiers of coverage -- such as preferred generic, other generic, preferred brand, other brand, and "specialty" -- with different copays or coinsurance at each tier, each of which can also vary by retail versus mail order, and for "preferred pharmacies."
Moreover, the health plan's formulary, which determines what tier a drug is in, can change at any time. Plus, as has been illustrated recently, the prices of any specific drug can change without notice, sometimes dramatically. If either of those happens to one of your drugs, say goodbye to your budget.
It's all enough to make your head spin.
The health plans would no doubt argue that their various approaches to prescription drug coverage are necessary in their efforts to control ever-rising costs for prescription drug costs. Well, they aren't working.
Prescription drug prices continue to soar, even for generic drugs. They have become a political issue, with the Senate now launching a bipartisan investigation into prescription drug pricing and the Presidential hopefuls from both parties being forced to take positions on how they would control them. For once, politicians are in sync with their constituents; the latest Kaiser Health Tracking Poll found that affordability of prescription drugs tops their priority list for Congress and the President.
I've long thought that the pharmaceutical industry was ahead of the rest of the health care industry. They were doing electronic submission of claims over forty years ago. They pushed for direct-to-consumer advertising in the late 1980's, and quickly jumped on that bandwagon. While providers only grudgingly adopted EHRs, they quickly moved to e-prescribing. Other health providers had to move away from discounted charges twenty years ago, whereas drug companies still mostly use that approach and are only starting to tip-toe into more "value-based" approaches, as with the recent Harvard Pilgrim-Amgen deal.
And the backroom rebate deals between drug manufacturers and payors put a lie to any claim that at least drug pricing is transparent.
It's not only prescription drug coverage that is increasingly complicated, what with narrow networks, gatekeepers, different copays for different types of medical services, bundled pricing, or numerous other gimmicks used in health plan designs. The collateral damage in the ongoing payor-provider arms race is consumer understanding.
Making things more complicated for consumers is not the answer.
In typical fashion, the health care industry has tried to address the confusion by creating a new industry that doesn't actually solve the problem but does manage to introduce new costs. Many enrollment sites --the Medicare plan finder, public exchanges, private exchanges, broker sites like ehealth, or health insurer sites -- offer tools that purport to estimate your costs under your various health plan options. Yet consumers still don't understand their options.
We keep treating health care as a multi-party arrangement between providers/health plans/employers/government/consumers, which is why everything ends up so complicated. Drug company rebates or medical device manufacturers' payments to providers are prime examples of the kind of insider trading that goes on. It's usually the consumers that come last. And that's the problem.
I think back to 1990's cell phone plans. Consumers never knew what their next bill would bring, between peak/non-peak minutes and the infamous roaming charges. No one liked it, no one understood it, and for several years no one did anything about it. Then AT&T came out with a flat rate plan that essentially said, "we'll worry about all those for you," and soon all carriers had to adopt a version of it.
I keep hoping for that kind of breakthrough with health insurance.
I'm torn between thinking that the people designing the plans are extremely clever, have a perverse sense of humor, or were under the influence of psychedelic drugs at the time.
It's not that there aren't plenty of options. I've got different levels of HMO, POS, and PPO options, from multiple carriers. My mother has many choices of Medicare Supplements, with Part D options, as well as Medicare Advantage options (both HMO and PPO), each from multiple health insurers.
Nor is it that having choices is bad. Researchers have discussed the "tyranny of choice" (or "paradox of choice") for some time, meaning too many choices can be paralyzing for consumers. I have to admit that when I go to the cereal aisle in the grocery store I feel overwhelmed, but, whether it is cereal or health plans, I'd still rather than more choices than fewer choices.
It's just that, well, the options are so damn confusing. I was in the health plan business for a long time, and helped develop some of the first plan selection tools for consumers, But when it comes to evaluating some of the options now available, I find it practically impossible.
Austin Frakt recently wrote in The New York Times about this problem. He cited a few studies specifically on point about health insurance, such as:
- One study found that 71% of consumers couldn't identify basic cost-sharing features;
- Less than a third of consumers in another study could correctly answer questions about their current coverage;
- Researchers found that consumers tended to choose plans labeled "gold" -- even when the researchers switched the "gold" and "bronze" designations, keeping all other plan details the same.
I'm most frustrated with prescription drug coverage. Not that long ago, the only variables were the copays for generic versus brand drugs. Now there are often five or six different tiers of coverage -- such as preferred generic, other generic, preferred brand, other brand, and "specialty" -- with different copays or coinsurance at each tier, each of which can also vary by retail versus mail order, and for "preferred pharmacies."
Moreover, the health plan's formulary, which determines what tier a drug is in, can change at any time. Plus, as has been illustrated recently, the prices of any specific drug can change without notice, sometimes dramatically. If either of those happens to one of your drugs, say goodbye to your budget.
It's all enough to make your head spin.
The health plans would no doubt argue that their various approaches to prescription drug coverage are necessary in their efforts to control ever-rising costs for prescription drug costs. Well, they aren't working.
Prescription drug prices continue to soar, even for generic drugs. They have become a political issue, with the Senate now launching a bipartisan investigation into prescription drug pricing and the Presidential hopefuls from both parties being forced to take positions on how they would control them. For once, politicians are in sync with their constituents; the latest Kaiser Health Tracking Poll found that affordability of prescription drugs tops their priority list for Congress and the President.
I've long thought that the pharmaceutical industry was ahead of the rest of the health care industry. They were doing electronic submission of claims over forty years ago. They pushed for direct-to-consumer advertising in the late 1980's, and quickly jumped on that bandwagon. While providers only grudgingly adopted EHRs, they quickly moved to e-prescribing. Other health providers had to move away from discounted charges twenty years ago, whereas drug companies still mostly use that approach and are only starting to tip-toe into more "value-based" approaches, as with the recent Harvard Pilgrim-Amgen deal.
And the backroom rebate deals between drug manufacturers and payors put a lie to any claim that at least drug pricing is transparent.
It's not only prescription drug coverage that is increasingly complicated, what with narrow networks, gatekeepers, different copays for different types of medical services, bundled pricing, or numerous other gimmicks used in health plan designs. The collateral damage in the ongoing payor-provider arms race is consumer understanding.
Making things more complicated for consumers is not the answer.
In typical fashion, the health care industry has tried to address the confusion by creating a new industry that doesn't actually solve the problem but does manage to introduce new costs. Many enrollment sites --the Medicare plan finder, public exchanges, private exchanges, broker sites like ehealth, or health insurer sites -- offer tools that purport to estimate your costs under your various health plan options. Yet consumers still don't understand their options.
We keep treating health care as a multi-party arrangement between providers/health plans/employers/government/consumers, which is why everything ends up so complicated. Drug company rebates or medical device manufacturers' payments to providers are prime examples of the kind of insider trading that goes on. It's usually the consumers that come last. And that's the problem.
I think back to 1990's cell phone plans. Consumers never knew what their next bill would bring, between peak/non-peak minutes and the infamous roaming charges. No one liked it, no one understood it, and for several years no one did anything about it. Then AT&T came out with a flat rate plan that essentially said, "we'll worry about all those for you," and soon all carriers had to adopt a version of it.
I keep hoping for that kind of breakthrough with health insurance.
Wednesday, November 4, 2015
My Phone Says I've Looked Better
I admit it; I'm a sucker for artificial intelligence. For example, this week Google said it has developed "Smart Reply," which uses AI to suggest potential replies to your emails, based on their contents, your previous replies, and other emails Gmail has spied on read. The responses are brief, but you have to assume the feature will only get smarter.
Meanwhile, Facebook says the facial recognition software it uses to help users tag photos -- which they claim is already almost as accurate as a human's -- can now recognize not only people in pictures but objects in the pictures with them. It can even answer questions about what is in the photos.
Of course, if Google's auto-correct suggestions or Facebook's success in tagging my photos are any indication, AI is still has a ways to go.
Despite that, experts are excited about the potential application of AI to health care, as evidenced by some speakers at last week's Connected Health Symposium, hosted by Partners HealthCare:
We will someday have AI-based clinical support systems to guide physicians in real time interactions with patients. We will someday have our own AI-based health coaches, integrated with our Cortana/Google Now/Siri/M interfaces on our various devices. And we will someday have our own AI physician avatars, supplementing or, in some cases, replacing our need for actual physicians.
Right now, though, I'm thinking about that facial recognition AI.
Current AI can sift through millions of photos to pick you out of a crowd, with varying degrees of success. Camera angles, make-up, hats, quality of image all factor into how successful such software is. Given the recent rapid rates of improvement, though, these are bumps in the road, not insurmountable barriers.
Other software can process your facial expressions, allowing them to make some good guesses about your emotions. If you are a marketer, or a law enforcement officer, this information might be gold, but if your privacy is important, it might be a scary invasion. Someone is always watching.
What I want to know is when this AI can tell if I look sick.
It can already, I assume, look at a set of pictures featuring me in varying degrees of illness and still determine if that the images are all, in fact, me. It would not seem much of a stretch for such software to look at a picture of me and determine if I don't look "normal." E.g., I'm pale, I'm in pain, or I'm unusually tired. Jaundice or measles would seem to be a piece of cake. Issues like a swollen ankle or a gash would also be obvious, as would a limp or rapid breathing, if a video or sequential still images are available.
Facial recognition software has had to learn to ignore certain variations from the norm -- e.g., a frown versus a smile, standing versus sitting --in a person's image, so this would amount to learning which variations are inconsequential and which may be attributable to a physical problem, like an illness or a severe injury. The next step would be to distinguish what that physical problem might be.
AI can already predict how you'll look as you age; it wouldn't seem hard to use similar AI to detect premature aging that could be linked to an illness -- or to visually illustrate longer term impacts of bad health habits. Even young "invincibles" might pay attention to the latter.
Physicians stress the importance of face-to-face visits with patients, claiming that they can pick up clues that they might miss over the phone or even via a video visit. AI should eventually be able to pick up on at least some of those clues.
Research indicates that smartphone users check their phones an astonishing 83 times a day (and the rate for younger users is some 50% higher than that!). That presents an easy opportunity for your phone to check on you and spot early warning signs.
One can easily imagine an app that, say, takes a photo of you at times throughout the day, perhaps periodically prompting you to note how you are feeling so it can better learn what your variations mean. The easy first stage would be for it to recognize variations that are unusual, even if the AI doesn't quite know what they might mean. That might warrant a "hey, maybe you should call your doctor," or a "GO TO THE ER NOW!"
The second, and harder stage, would be to match your variations with how selected conditions visually present. For example, it could recognize a rash or a growth, and maybe effects of a stroke. One can imagine an AI synthesizing large data-sets of photos of people with specific issues to derive some commonalities, and use that learning to suggest what you might be experiencing. The AI might need to ask you what you were feeling before suggesting a potential diagnosis, but at least it could help narrow what might be the problem -- if you even realized there was a problem.
I don't really mind writing my own emails and I'm happy to tag my own photos, thank you very much, but if an AI can help me recognize when I have a health issue, that'd be something I could use.
Meanwhile, Facebook says the facial recognition software it uses to help users tag photos -- which they claim is already almost as accurate as a human's -- can now recognize not only people in pictures but objects in the pictures with them. It can even answer questions about what is in the photos.
Of course, if Google's auto-correct suggestions or Facebook's success in tagging my photos are any indication, AI is still has a ways to go.
Despite that, experts are excited about the potential application of AI to health care, as evidenced by some speakers at last week's Connected Health Symposium, hosted by Partners HealthCare:
- Partner's Dr. Joseph Kvedar foresees us having automated health coaches in our smartphones. "This is quite doable,” he said. "It’s as if the puzzle pieces are there and we haven’t put the jigsaw puzzle together yet.”
- MIT's Joi Ito sees the doctor-computer combination as "the winning combination." He used the now-familiar example of IBM's Watson, saying: "I think there was an announcement recently that Watson is almost finished with med school, It’s sort of a joke, but sort of true...Now imagine if you had a computer that had all of the knowledge that you needed for med school and if it were available all of the time, maybe there’s an argument to be made that you don’t have to memorize it if it’s available all of the time.”
- VC mogul Vinod Khosha predicts AI will eventually help us make better diagnoses. As he points out: "The error rates in medicine, if you look at the Institute of Medicine studies, are about the same as if Google’s self driving car was allowed to drive if it only had one accident per week." Ouch.
We will someday have AI-based clinical support systems to guide physicians in real time interactions with patients. We will someday have our own AI-based health coaches, integrated with our Cortana/Google Now/Siri/M interfaces on our various devices. And we will someday have our own AI physician avatars, supplementing or, in some cases, replacing our need for actual physicians.
Right now, though, I'm thinking about that facial recognition AI.
Current AI can sift through millions of photos to pick you out of a crowd, with varying degrees of success. Camera angles, make-up, hats, quality of image all factor into how successful such software is. Given the recent rapid rates of improvement, though, these are bumps in the road, not insurmountable barriers.
Other software can process your facial expressions, allowing them to make some good guesses about your emotions. If you are a marketer, or a law enforcement officer, this information might be gold, but if your privacy is important, it might be a scary invasion. Someone is always watching.
What I want to know is when this AI can tell if I look sick.
It can already, I assume, look at a set of pictures featuring me in varying degrees of illness and still determine if that the images are all, in fact, me. It would not seem much of a stretch for such software to look at a picture of me and determine if I don't look "normal." E.g., I'm pale, I'm in pain, or I'm unusually tired. Jaundice or measles would seem to be a piece of cake. Issues like a swollen ankle or a gash would also be obvious, as would a limp or rapid breathing, if a video or sequential still images are available.
Facial recognition software has had to learn to ignore certain variations from the norm -- e.g., a frown versus a smile, standing versus sitting --in a person's image, so this would amount to learning which variations are inconsequential and which may be attributable to a physical problem, like an illness or a severe injury. The next step would be to distinguish what that physical problem might be.
AI can already predict how you'll look as you age; it wouldn't seem hard to use similar AI to detect premature aging that could be linked to an illness -- or to visually illustrate longer term impacts of bad health habits. Even young "invincibles" might pay attention to the latter.
Physicians stress the importance of face-to-face visits with patients, claiming that they can pick up clues that they might miss over the phone or even via a video visit. AI should eventually be able to pick up on at least some of those clues.
Research indicates that smartphone users check their phones an astonishing 83 times a day (and the rate for younger users is some 50% higher than that!). That presents an easy opportunity for your phone to check on you and spot early warning signs.
One can easily imagine an app that, say, takes a photo of you at times throughout the day, perhaps periodically prompting you to note how you are feeling so it can better learn what your variations mean. The easy first stage would be for it to recognize variations that are unusual, even if the AI doesn't quite know what they might mean. That might warrant a "hey, maybe you should call your doctor," or a "GO TO THE ER NOW!"
The second, and harder stage, would be to match your variations with how selected conditions visually present. For example, it could recognize a rash or a growth, and maybe effects of a stroke. One can imagine an AI synthesizing large data-sets of photos of people with specific issues to derive some commonalities, and use that learning to suggest what you might be experiencing. The AI might need to ask you what you were feeling before suggesting a potential diagnosis, but at least it could help narrow what might be the problem -- if you even realized there was a problem.
I don't really mind writing my own emails and I'm happy to tag my own photos, thank you very much, but if an AI can help me recognize when I have a health issue, that'd be something I could use.
Subscribe to:
Posts (Atom)