Thank you for this fascinating video! I'm doing my psychology honours study in a cognitive neuroscience topic, but your channel has made me extremely interested in computational approaches which I haven't had any exposure to in my degree. I can see the potential benefit of studying certain disorders using something like bayesian inference. E.g. in anxiety, where it intuitively makes sense to me that prior beliefs and biases have an unusually strong impact on present experiences. Going to read the papers you mentioned in the video :)
I think equations can be used in psychiatry, but what we have to keep in mind can be conveyed through this quote: "All models are wrong, some models are useful". We shall use models for some insights but we need to be cautious about how much we are driving our decisions from the model inferences. I truely resonate with your opinions and many thanks for the material you have shared here. I am a 3rd year Neuroscience PhD student from India.
For sure equations/modelling could provide answers provided they are falsifiable through experiments, since the brain is fundamentally a physical system. However it's very easy for someone to come up with some complicated equations to back up a theory they might have, giving it an air of legitimacy, but without solid experimental backing this is just a hypothesis or more like philosophy. There's nothing wrong with writing papers like that, and they can be very interesting, but I think they should be treated like philosophy rather than solid science that can be taken as fact. I think this is perhaps where the controversy can come from - where authors or others claim this kind of thing is anything more than philosophy whilst having shaky or no experimental backing, or worse - no forseeable way of performing experiments that would actually rigourously test their theory.
Hey Charlotte ! Thank you for all these recommandations. Thanks to you, I'm currently following the Zurich Computational Psychiatry Course and it's really interesting. Your speech this morning was really great and helpful.
On your question to the audience: I think that trying to computationally and equationally calculate mental illnesses is an absolutely workable idea, since I believe that there are no areas “too human/animate” to try to calculate them, but only insufficient knowledge of the area or limitations in computing power/time. Thank you for the video)) I enjoy the quotes you choose! Thank you for the time you spend to choose them from your first and second brains))
Oh my goodness this is an amazing collection of information - I did not realize how much of my own research and work is connected to this field! Thanks so much for sharing all of this, very much looking forward to learning more :)
Would love to hear what channels on youtube you've found that are related to your research, I'm on a quest to make an index of academic youtube :D Any chance you'd be willing to make your subscriptions public? Researcher subscriptions often have very interesting channel finds in them.
You are doing all the things that I ever wanted to do for answer my questions in my field. Sadly, hear I don't have to much tools to get properly form in translational psychiatry. Love you're videos!
Thank you for the video, please keep making more! As a current Neuro master student looking for the areas of interest you provided me with so much inspiration and knowledge! ❤
OMG! i think I found something I love! Thank you for exposing this topic. I have a degree in Neuroscience but was unsure how to mold my passion for philosophy, psychology, and decision making theories! I think I would love to study computational psychiatry! I’m looking to get a masters but not sure where to begin.
You could think of the dopamine circuit for example in those with depression being dyresgulated and running trials to see how individual responses to rewards correspond to the firing of neurons as well as gaining biological markers and psychological assessments to accelerate this field. Certainly an effort that requires many brilliant minds! Bayesian models seem very exciting and promising if we can ever get to see prediction coding as it happens in real time. Seems out of this world.
Thanks for the video and for the papers! I think an interesting question that rises here is that the models in computational psychiatry should be interpretable. So I would argue that using neural networks in psychiatry in practice might be highly controversial.
Hi, Charlotte, this video could not have come at a better time for me, since I just started reading the first article you recommended and I was struggling to find more sources :), thank you! Apart from all these factors you mentioned, I think that the evolutionary perspective is also very interesting. If mental disorders are regarded as undesirable traits, how come they were not eliminated during the course of evolution? I think that perhaps natural responses to physical danger (fear) have been replaced by anxiety as our world became safer, but also more complicated. I agree that there is a great deal of controversy regarding all these things.
So interesting thank you so much for this - definitely going to look at those references and that online course looks great. As a regular psychiatrist (not a computational one), do I think computational methods should be used to look at and further our understanding of the brain-mind? Yes. However, I'm not sure applying them to the categories of mental illness as defined by DSM/ICD is going to get us very far because of the heterogeneity within diagnoses, overlap between diagnoses (and the many other flaws with diagnosis I'm sure you're aware of). I do hope it will further our understanding of the brain-mind though, and therefore in time will improve our understanding of the things that the brain-mind--body does that causes people distress and problems.
hi! just discovered your channel about 10 minutes ago. is there anything about your career/academic path to computational psychiatry? thank you so much!!
From my perspective I think this computational approach could stop overdiagnosing or miss diagnosing along with a proper treatment, modeling the propected mental illness and finding a proper theraphy that fits the person diagnosis would help us get the subject off the medication much sooner.
This is a terribly wrong comment, and I will answer you with two quotes from the Math 51 course text prepared by the Stanford University Math Department (Linear Algebra, Multivariable Calculus, and Modern Applications)... “Linear algebra is the central subject of mathematics. You can’t learn too much linear algebra.” Benedict Gross, former Dean of Harvard College “The revolution of the 21st century [is] going to be to make medicine a computational science.” Murray Sachs, pioneering scientist, biomedical engineer
