Someone please make an app that makes psychiatric analysis based on voice! I can think of a million uses, from tracking my own mood to secretly diagnosing my co-workers.
The benefits of this when implemented correctly nonwithstanding, there's a major problem in that any implementation is highly likely to be quite flawed. See the proliferation of "health-tracking" apps that are basically quackery. With psychiatric analysis --- where the signals to track are much weakly defined than heart rate, blood pressure, etc. --- it is far too likely that apps will consistently make poor diagnoses, to the detriment of the user.
I did plenty of graduate work with speech recognition; and detecting cognitive states based on speech is not much better than a coin-flip. Depression, drunkenness, deceit, happiness, etc. are hard to detect for an arbitrary person with accuracy. Speech signals for latent states in human behavior are highly idiosyncratic. The answer I always give when asked if something can be detected by speech (most requested - lying): "Yes, but only if you have a lot of time and want it to be accurate for at most a couple people. And it's going to be expensive."
Agreed that this alone would be insufficient but there's value in sharing this information with the patient's doctor, who could choose to ignore it if there's no other evidence.
Besides, we have to start somewhere, right? If I had the opportunity, I would release it as an open-source project for other medical professionals to improve upon. I certainly wouldn't go as far as submitting it to the app store for the reasons you mentioned, i.e. the detriment to the user.
Just seemed weird to me to mention doing it in a secretive/unconsenting manner, of people you're supposed to trust, and handing off private conversations to third parties without asking. It says a lot about respect toward one another.
I think that part was sarcastic. Everyone hates that guy who reads five sentences from Wikipedia and is convinced you have an especially bad case of the Bajoran flu.
It'd take a lot of speech data to make a diagnosis, and I'd imagine it'd be error-prone at first.
As for how to differentiate hypomania from excitement, I'm skeptical that it can be done. But if someone who usually talks 50% of the time in conversations is taking 95% of the time, and this persists over weeks, there's a signal there. Also, hypomania tends to produce a lot of run-on and deeply-nested (even Lispy (as in having multiple levels of implicit parentheses)) sentences.
Not enough personal experience with actual mania (as opposed to hypomania) to opine on it or guess what its signals might be. Only had it a couple times and none recently.
I totally need this...having suffered from several manic episodes that usually ramp up over a period of a month or two before I become straight up psychotic, completely lost. It would change my life to be able to check in with my psychiatrist after several alerts, rather than when I get to the point that I need to be admitted. its also not the type of message that friends and family like to give.
oh man the feels, stay strong friend. mental illness history myself. gotta remember to stay safe and make sure others don't see you as a threat. try to do things that make your life calmer and more stress free. it's the only way to counter the mood swings.
Agreed, but it also narrows down on certain things. For e.g. it could mean that the computing power required is small, and that you don't need an expensive microphone that captures a wide frequency range.
I totally get what your saying about it being link baitish. However,if the manic compression in my voice could be caught by my phone before I make an urgent call, it would be most advantageous, as opposed to a microphone on my computer.
you have to remember that the only microphones people have any exposure to in regular life is their phone. So it makes sense that any technology that uses a microphone be found in that device.
It's not even linkbait because it's overly connected to cell phones.
It's linkbait b/c it's not as certain as the title says.
TFA says, emphasis mine:
features like pitch, loudness, and tempo that may predict a mood swing in the near future, before any recognizable change in mood emerges. For example, a single feature, like loudness, may flag a person on the path to mania or depression once it crosses some threshold. Or telltale signs could come from a combination of these features, or the way they change over a few days. The project was funded in 2013 by the National Institute of Mental Health, so it’s still in its early days.Though they don’t yet know what will make the best predictors,
google likely already know from your searches for Greek real estate and Ferrari cars. I'm sure search history analysis can reveal depression just as effectively.
People, please don't show a continuous variable with a rainbow gradient. Two colors will do - red/green, green/black, red/blue. Too many colors and you lose reference. You see green and have to think about what part of the distribution it maps to; this defeats the whole purpose. Rainbows are pretty, but not the best choice.
28 comments
[ 2.9 ms ] story [ 78.7 ms ] threadThe benefits of this when implemented correctly nonwithstanding, there's a major problem in that any implementation is highly likely to be quite flawed. See the proliferation of "health-tracking" apps that are basically quackery. With psychiatric analysis --- where the signals to track are much weakly defined than heart rate, blood pressure, etc. --- it is far too likely that apps will consistently make poor diagnoses, to the detriment of the user.
Besides, we have to start somewhere, right? If I had the opportunity, I would release it as an open-source project for other medical professionals to improve upon. I certainly wouldn't go as far as submitting it to the app store for the reasons you mentioned, i.e. the detriment to the user.
Interesting that you explicitly mention you'd do it without their knowledge/consent...
http://www.nytimes.com/2012/06/17/opinion/sunday/how-depress...
As for how to differentiate hypomania from excitement, I'm skeptical that it can be done. But if someone who usually talks 50% of the time in conversations is taking 95% of the time, and this persists over weeks, there's a signal there. Also, hypomania tends to produce a lot of run-on and deeply-nested (even Lispy (as in having multiple levels of implicit parentheses)) sentences.
Not enough personal experience with actual mania (as opposed to hypomania) to opine on it or guess what its signals might be. Only had it a couple times and none recently.
It's linkbait b/c it's not as certain as the title says.
TFA says, emphasis mine:
features like pitch, loudness, and tempo that may predict a mood swing in the near future, before any recognizable change in mood emerges. For example, a single feature, like loudness, may flag a person on the path to mania or depression once it crosses some threshold. Or telltale signs could come from a combination of these features, or the way they change over a few days. The project was funded in 2013 by the National Institute of Mental Health, so it’s still in its early days. Though they don’t yet know what will make the best predictors,