This is the rather more interesting part of the article; not just claiming some magic task AI can do, but explaining how it does it, and what it bases its decisions on. This means that in theory, someone can validate whether this approach is actually true, or if (as others have pointed out) the AI is just reading something else off those images that we're not aware of.
> Until recently, a model like the one Menon’s team employed would help researchers sort brains into different groups but wouldn’t provide information about how the sorting happened. Today, however, researchers have access to a tool called “explainable AI,” which can sift through vast amounts of data to explain how a model’s decisions are made.
> Using explainable AI, Menon and his team identified the brain networks that were most important to the model’s judgment of whether a brain scan came from a man or a woman. They found the model was most often looking to the default mode network, striatum, and the limbic network to make the call.
This (feature or attention maps) is BS. A real explanation would show what factors were being used to make the determination. Claiming the model is "looking at" something has no explanatory power. We already knew it was looking at the picture. These heat maps are great for showing people what they want to see and making wishy washy claims like this but they really don't explain anything in the commonly understood sense of telling us what is different.
I disagree. It clearly tells us _what_ is different, it just doesn't tell us _how_ it differs from case to case. So it's not a full explanation (which would, as you note, require a model of how it differs and, optimally, _why_) but it is a step towards a explanation, and not to be sneered at.
I might get decapitated for suggesting this here, but it would be interesting to see the results on people with gender dysphoria. Do the hormones change the brain or does it remain the same? Test it both on patients starting as teens and as adults and study over time.
If I can derail conversation for a moment, what is the idea behind 'valid' or 'invalid' transition? It's not a phrase I've heard before sorry. I'm not in any kind of other forum where I think I could ask this without getting very emotionally-charged responses.
Ah that makes sense. I think I took it a little too literally and the phrasing lead me to think that some transgender people themselves categorise transitioning as valid or invalid.
Look into gender essentialism, which contrasts with gender constructivism.
It's often associated with TERFs, but it is VERY common within transgender communities themselves. There's notion that certain gender identities are valid and others are not (especially being male/female), and there's also the notion that diagnosable medical issues are required to be truly transgender.
In this conversation, if you're AMAB, and have a male brain, and claim to have transitioned to being non-binary a gender essentialist might go "aha! Science proves they are nothing but a transtrender and a man" A constructionist would say "aha - the fact they have a male brain yet expresses a non-male gender identity proves they're a truly unique and special transgender person".
It seems like an easy way to be on the "Right side of history" to reject gender essentialism, since you end up never calling a transgender person invalid. There is a dark side to rejecting essentialism, which is if gender ISN'T essential, than logically it should be possible to change people's gender identity to a "better" one. Which can result in things like Joan/John https://en.wikipedia.org/wiki/David_Reimer.
In this case I meant valid as in, transitioning is really the best way to treat gender dysphoria and it's not useful to think of dysphoria as a delusion that would go away with conversion therapy or with time.
There have been studies that have shown that a persons preference for things over interactions were affected by testosterone levels from a young age (1 or 2) to teenage years. That would also affect the brain development etc..
That being the case, unless you changed it at a young age I'm not sure that this could be changed later on in life.
Many people suggest that trans people have brains more inline with their desired gender, and this causes feelings of gender incongruence.
However, if we take a closer look, we see some important underlying issues.
In a famous study, the authors compared prepubertal and adolescent children, then suggest sex atypical cerebral differentiation occurs within these individuals [1].
The authors found sex atypical differences in the adolescent cohort, but the majority of that cohort is homosexual:
- Homosexuality = 23% of trans boys + 44% of trans girls (prepubertal cohort)
- Homosexuality = 100% of trans boys + 78% of trans girls (adolescent cohort)
The only non-sex typical finding which was specific to gender dysphoria was in visual network-1 (VN-1; via fMRI). It was suggested that alterations in this network may disrupt body perception in gender dysphoric individuals.
In another study by some of the same authors, they tested whether transgender people (with gender dysphoria) would have sex atypical hypothalamic activation to androstenedione, a steroid hormone in human sweat that causes sex-specific olfactory responses [2].
But similar to the previous study, sexual orientation was not accounted for. Why is this important? Well, the same sex atypical hypothalamic response is observed in homosexual men [3] and in homosexual women [4].
It's becoming increasingly clear that the only people with any sort of sex atypical cerebral differentiation occurs in homosexual individuals (on average).
This is further supported by functional connectivity studies of sex atypical amygdala co-variance. You can see this in Figure 1 of this paper: [5]. Notice the high similarity in amygdala activity (at rest) between heterosexual females and homosexual males (and vice versa).
Interestingly, in an effort to bring all of this together, this study examined the brains of heterosexual transgender people in order to control for sexual orientation: [6]. As Figure 1 of this paper shows, the authors found sexual dimorphism in various gray matter parameters in control male and females. However, these findings were not found in the heterosexual transgender population.
So rather than sex atypical brain structure/function, what is specific to gender dysphoria itself?
It's been shown that individuals with gender dysphoria show weaker structural and functional connectivity within the default mode network (DMN) of the brain, which is vital for body perception/image and self-referential processing [7]. Findings which have been replicated in [8] and [9].
The DMN consists of cerebral midline structures, including the medial prefrontal cortex (mPFC) and the posterior cingulate cortex (PCC). Interestingly, it's been shown that trans individuals (with GD) show a stronger activation pattern within these DMN structures when viewing pictures of their body morphed to the opposite sex: [10]. You can see the results in Figure 5 of this paper.
It's important to note that correlation is not causation. Just because we observe a different pattern in gender dysphoric subjects compared to neuro-typical controls, does not suggest it's innate (born with) or a product of post-natal experience. We simply do not know.
Also, transgender people tend to have lots of co-morbidities from depression, anxiety, anorexia, autism, and a homosexual orientation. All of this needs to be considered when looking at neuroscience studies on this population.
[1] Nota, N. M., Kreukels, B. P. C., den Heijer, M., Veltman, D. J., Cohen-Kettenis, P. T., Burke, S. M., & Bakker, J. (2017). Brain functional connectivity patterns in children and adolescents with gender dysphoria: Sex-atypical or not? https://pubmed.ncbi.nlm.nih.gov/28972892/
[2] Burke, S. M., Cohen-Kettenis, P. T., Veltman, D. J., Klink, D. T., & Bakker, J. (2014). Hypothalamic response to the chemo-signal androstadienone in gender dysphoric children and adolescents. https://pubmed.ncbi.nlm.nih.gov/24904525/
[5] Savic, I., & Lindström, P. (2008). PET and MRI show differences in cerebral asymmetry and functional connectivity between homo- and heterosexual subjects. https://pubmed.ncbi.nlm.nih.gov/18559854/
[7] Burke, S. M., Manzouri, A. H., & Savic, I. (2017). Structural connections in the brain in relation to gender identity and sexual orientation. https://pubmed.ncbi.nlm.nih.gov/29263327/
[8] Uribe, C., Junque, C., Gómez-Gil, E., Abos, A., Mueller, S. C., & Guillamon, A. (2020). Brain network interactions in transgender individuals with gender incongruence. https://pubmed.ncbi.nlm.nih.gov/32057995/
[9] Feusner, J. D., Lidström, A., Moody, T. D., Dhejne, C., Bookheimer, S. Y., & Savic, I. (2017). Intrinsic network connectivity and own body perception in gender dysphoria. https://pubmed.ncbi.nlm.nih.gov/27444730/
[10] Majid, D. S. A., Burke, S. M., Manzouri, A., Moody, T. D., Dhejne, C., Feusner, J. D., & Savic, I. (2020). Neural Systems for Own-body Processing Align with Gender Identity Rather Than Birth-assigned Sex. https://pubmed.ncbi.nlm.nih.gov/31813993/
> This breakthrough supports the theory that significant sex differences in brain organization exist, challenging long-standing controversies.
It doesn't, because there is no explanation for how it determines the difference. It's just as likely some confounding thing that's not causally based on the brain, like all these other "AI can tell" studies that don't identify an actual mechanism.
That said I'm surprised that we can't tell men and women apart from the brain. I would have assumed there would be some differences that would make it obvious at least for the average brain. It's more surprising to find they are indistinguishable.
>It doesn't, because there is no explanation for how it determines the difference.
You don't need to know the mechanism to prove that it "supports the theory that there is a difference in brain organization". If there wasn't a difference, it wouldn't be able to find it.
>It's just as likely some confounding thing that's not causally based on the brain
Given that it does it's identification by looking at the brain, whatever the cause is, it is ALSO manifested as a "difference in brain organization".
>That said I'm surprised that we can't tell men and women apart from the brain.
Who said we can't? The article points to that we can.
And there are other brain attributes besides organization we can use to tell them apart, the biggest being size, including size of certain structures.
> Given that it does it's identification by looking at the brain, whatever the cause is, it is ALSO manifested as a "difference in brain organization".
I think the person you’re responding to is talking about confounding factors _outside_ the brain scan, like when a model that could identify skin cancer used the presence of a ruler as its most weighted input: https://venturebeat.com/business/when-ai-flags-the-ruler-not...
This is a huge problem in medical use for AI - you have non-technical medical professionals influencing the data used to train the model, and introducing biases in the data that the model is learning from.
>I think the person you’re responding to is talking about confounding factors _outside_ the brain scan, like when a model that could identify skin cancer used the presence of a ruler as its most weighted input
Well, those are just "bugs" and the results are then useless.
My point is that results taking into account only the scan (not random correlations with outside factors), would indeed be able to prove differences between M&F brains, without needing to also give a mechanism for it.
Some silly examples of stuff it might find that aren't actually brain differences:
1. There's an M/F letter or other identifying text on the scans
2. Men and women get put into the machine slightly differently, and the ai sees a slight difference in the pose of the scan
3. It sees something else that's identifying and pretty good at splitting, like ear holes for ear rings
I remember reading about one of the skin cancer finding ML models that was developed. Most of the pictures from malignant moles were taken from a specific office and the algorithm was picking up on the lighting in that room rather than other differences in the mole to determine likelihood of it being cancerous. It did a great job of identifying photos based on the type of lighting used, but was overall less effective at predicting cancer.
I’m reminded of the tale of a similar breakthrough, and upon ablation, it was found to be indexing on a particular artifact found on one X-ray machine. Or maybe the patient name?
Anyways, would recommend some caution, there’s no reason to believe there’s some obvious fundamental difference in brain organization that we can’t see ourselves, until the research is a bit further along :)
Unfortunately, these things can be extremely subtle. Far too often it boils down to sampling bias where the differences are actually differences in the variance estimation that results in a significant result. For example suppose that "male" dice are biased to role a 5 slightly more often than "female" dice. You will find a statistically significant difference related to the 5-side of the dice if you roll them enough. It doesn't mean that rolling a 5 indicates anything about the "gender" of a particular dice.
> It doesn't, because there is no explanation for how it determines the difference
I don’t think an explanation is required for this.
I should be able to state that two species have a different lifespan based on a difference in mean age of death - without needing to provide any specific biological mechanism in my study.
Substitute that mean for a linear regression / neural network and the species lifespan for some vector of brain organisation to get the results in the article.
I was thinking about this too, but I don’t know how widely they overlap? Like, if (say) 30% of male brains are at least 10% smaller than the average male brain, and 30% of female brains are at least 10% larger than the average female brain, then a model that just considered volume might not be very accurate at all, because anything in the middle still has a significant chance of being either sex?
> I should be able to state that two species have a different lifespan based on a difference in mean age of death - without needing to provide any specific biological mechanism in my study.
You need to control for confounders. For example, if I told you:
- Mean lifespan of species 1 is: 3 years
- Mean lifespan of species 2 is: 10 years
Does that mean that species 2 has a longer lifespan than species 1?
What happens when you find out that species 1 is domesticated cattle, which has (almost) all of its males slaughtered at age 2?
You could then say "I meant for two species that are in controlled conditions", but now you need to define what conditions you're controlling, which implies what casual mechanisms you're controlling for.
> which implies what casual mechanisms you're controlling for.
Without thinking terribly deeply about it, it seems like this could be modified to
"what correlated mechanisms we're controlling for", in which case, yes I think if we controlled for the highly correlated variable of "does this cow live on a meat farm" with the mean age, we would be able to make broad claims about cattle lifespan without needing to be privy to the internal mechanisms of the agriculture industry
> Until recently, a model like the one Menon’s team employed would help researchers sort brains into different groups but wouldn’t provide information about how the sorting happened. Today, however, researchers have access to a tool called “explainable AI,” which can sift through vast amounts of data to explain how a model’s decisions are made.
> Using explainable AI, Menon and his team identified the brain networks that were most important to the model’s judgment of whether a brain scan came from a man or a woman. They found the model was most often looking to the default mode network, striatum, and the limbic network to make the call.
What's more fascinating it that it's not just a parlor trick and actually can be "validated" in some sense by doing a cognition test for just that part of the brain and observing the differences. So the model said "check here and you'll see a difference," and apparently, they did.
> They developed sex-specific models of cognitive abilities: One model effectively predicted cognitive performance in men but not women, and another in women but not men. The findings indicate that functional brain characteristics varying between sexes have significant behavioral implications.
> “These models worked really well because we successfully separated brain patterns between sexes,” Menon said. “That tells me that overlooking sex differences in brain organization could lead us to miss key factors underlying neuropsychiatric disorders.”
FWIW, feature maps are not really reliable (to put it mildly). Knowing what area the model is "looking at" is virtually meaningless if you don't know what it's looking at. We already knew it was looking at the brain so is it turtles all the way down? Those maps are great for confirming what people want to see "oh look it tells a cat from a dog by the eyes" but have no explanatory power. If they did then the paper would be about AI finding a physiological difference and would tell us what it is, not some handwaving about "here's where it looked".
> Knowing what area the model is "looking at" is virtually meaningless if you don't know what it's looking at. We already knew it was looking at the brain so is it turtles all the way down?
Huh?
I’m not sure what you are saying here.
If there’s some subset of the pixels or whatever which the models predictions depend on more than the others, then, it depends on those more than it depends on the others.
Obviously it would be using the data it is given. That, doesn’t mean that looking into “what part of the data that it is given, is it using?” not meaningful.
> “These models worked really well because we successfully separated brain patterns between sexes,” Menon said. “That tells me that overlooking sex differences in brain organization could lead us to miss key factors underlying neuropsychiatric disorders.”
Although never demonstrated, it's always been most intuitive to me that sex-related brain differences underlie gender dysphoria. I would feel exceptionally relieved and vindicated to finally have some objective, falsifiable observation to point at that suggests the reality of dysphoria.
Ah, but this is science. How would you feel if it were not vindicated? What if it were not a matter of "born that way"?
Plus, there's a whole ton of confounding business we'll have to look out for. Just off the top of my head, you've got certainty intervals. And wouldn't it be prudent to at least examine the effects of various hormones like testosterone, dihydrotestosterone, estradiol, progesterone on the brain? Say, post age twenty five, pre age twenty five. And so on. Might exist, might not.
We're a long way from being able to have much certainty at all.
I would be shocked if hormones didn't have some effect on the brain considering how much of an effect they have everywhere else. Hell, I know some people who take the "brain sex" angle re. transness speculate that varying levels of hormonal exposure in utero could be what causes gender dysphoria post-utero... in which case, would adult endogenous hormones be responsible for brain structures, or would brain structure dictate which kind of hormones your body expects?
I think, in utero, the effects would be strong and likely visible, but through childhood, puberty, and then young adulthood, I would expect diminishing degrees of impact. Certainly, I've read case studies of various pre-natal exposures to particular chemicals having large and unmistakable impacts on sexual orientation. It's of especial interest to me as one of the very last DES babies.
Brain structure, as in the gross features visible on imaging, would probably not dictate what kinds of hormones you body might expect. I think that sort of thing would occur at a much lower level, right on the receptors of the tissues, and would not be visible.
I think it would be the same as right now, absence of evidence and all that. Would still be short one explanation explaining the observed phenomenon. You're absolutely right that it wouldn't be proof in any sense, just a sign that the avenue would be worth pursuing.
I love that idea! Honestly even if it doesn't show up as a sex-difference doing this same trick might be nonetheless fruitful by asking it to predict say cis woman / trans man.
> I believe this is a clue to where the GP post is coming from: we can't, therefore the study must be flawed.
How so? They are clearly saying they expected this to be possible already, until this article claimed otherwise (by suggesting AI is contributing something new here).
They are saying that this result will not resolve the disputes, because it doesn't address the core thing in dispute: that there are meaningful differences that people actually care about. For example, maybe it is possible to tell the sex based on the shape of the brain. This doesn't mean that men and woman think differently, which is what people are actually arguing about. And the AI in this study can't proof that, because it doesn't give us any further insights into how cognition works.
This point strikes me as actually pretty mundane and obviously correct. The fact that 5 people immediately seem to have misunderstood it (as in, they are not responding to the argument) seems to tell us something about their priors instead.
>They are saying that this result will not resolve the disputes, because it doesn't address the core thing in dispute: that there are meaningful differences that people actually care about.
That's not their claim. Their claim is that there is some flaw ("confounding thing") in the study. They then go on to say that men and women's brains are "indistinguishable." Well, of course you will automatically assume there is a flaw if that is your belief.
There’s some linear and obvious differences like the brain volume. Don’t need ai for that lol. I think I remember that it’s a 20% average difference. Downvote me please
This is comparable to trying to determine sex based on secondary sex characteristics.
Sex is binary, based on gametes which are binary. Anything with a bimodal distribution will by necessity never be the right tool for determining something which is truly binary.
The next best determinant of sex other than gametes is the presence or absence of the SRY gene.
I don’t like comments like this. I’m not picking on you in particular because I see them all the time, especially as drive-by comments. What I like a lot about HN is that typically these drive-by comments are not highly rewarded.
There is a kind of cocktail-party level expertise that is behind comments that scientific results are obvious. First of all, even if they were, it’s likely there’s a novel angle the researchers are targeting. We should ask ourselves why a researcher would spend time on such “obvious” results.
Conveniently, the article author asked the researchers why they would study this. It’s not just a question of whether their brains are dissimilar, but how.
Second, whether male and female brains are structurally similar is controversial. As the article states, brain structures tend to look mostly the same in men and women.
Forensic investigators can reconstruct the face from a skull, not only race/ethnicity/gender. It is probably just a question, how to measure improvement of AI, added to existing method. Let's say you find the skull of the alleged "new Saddam Hussein", the AI reconstructs the face, the face looks "similar", but are you willing to go along with the AI? Maybe the neural network just got good at drawing the face of the average Arab.
This is going to turn out to be the new phrenology, completely flawed, but it won't stop phobic types from using it to "prove" their antisocial theories.
I think the implication is it doesn't actually work well. Running hard-to-explain models on brain scans has a history of producing papers which fail to reproduce, there were a bunch of bunk results in fMRI research a few years ago.
While skepticism is healthy, I don't understand how you can say with certainty it will not work.
A hypothetical technology that could very accurately determine the sex of a brain would seem to be a boon for transgendered people. Early detection could mean better treatment and less suffering with fewer cases of desisting.
It depends on what the 90% is. The research is unpublished so there's no way to know.
What is the specificity and sensitivity? Or hell just give me a confusion matrix. That'll give you the answers you care about.
If the sensitivity is extremely high then it's an excellent diagnostic for sexual characteristics that are a deviation from the norm. But otherwise tbh I just don't see it being a super useful test.
Well if there's a high sensitivity in the test (which to get 0.9 would suggest a lower specificity) then that'd mean it always can tell when a brain deviates from the "established" sex.
i.e. for people who ahead of time aren't known to be intersex or trans, it would be able to identify them with high accuracy.
And the sensitivity vs specificity will depend for each category they are testing for (i.e. testing cis male, cis female, trans masc, trans fem, nonbinary, and some of the different intersex categories).
i.e It could be a cheaper test to do than genetic testing and if particularly effective may be useful in identifying genetics associated with different sex characteristics.
Usually, “sex” refers to biology (“what’s in your pants”) whereas “gender” refers to the social construct (are people expect you to wear pants?).
“Asexual” then usually refers to someone’s preferences about getting what’s in their pants and what’s in someone else’s pants together; asexual typically means “they’re not interested in that”.(“heterosexual” meaning they want what’s in the other person’s pants to be different than what’s in theirs)
There’s a lot of namespace collisions in all this.
> Usually, “sex” refers to biology (“what’s in your pants”)
"Sex" refers more specifically to hormonal expression than anything else.
What's in your pants, what chromosomes you have, etc aren't really great indicators for that. What hormones you have in your body in practice is really the only surefire thing.
This applies especially in the historical context since after social clues, the main thing you have left is osteology/bone analysis and that's driven predominantly by hormones more than anything else.
And even if you subtract the people on hormone replacement therapy, intersex people (people who naturally express "nonstandard" sex characteristics) make up upwards of 2% of the population.
So in the end sex really only refers to what your hormones are and nowadays you can change that with medication.
Sorry for being pedantic but this is one of the things that irks me with sex/gender discourse
While technically yes in the strict biological definition, it's a measurably less useful definition in humans since we generally don't consider a man who can't produce sperm as "not male" and a woman who doesn't have eggs as "not female".
Point being that while sex is certaintly more fundamental than gender, I don't know if the formal biology definition of sex is sufficiently nuanced for the way we perceive sex in humans. Especially since of the primary sex characteristics, having eggs or sperm is arguably the characteristic that humans care the least about when determining sex.
True, a male can develop with malformed testes or a woman with non-functioning ovaries, but can we still understand the difference between the sexes. For gonochoric species like humans it's the whole body plan that develops separately for each reproductive role.
It's not just hormone levels measured at any one point in time, but the result of a long and specialised developmental process. For instance, females and males end up having different skeletal structure which we can observe as accommodating childbirth in the former.
>So in the end sex really only refers to what your hormones are and nowadays you can change that with medication
I think there's a significant point being glossed over here. You are suggesting that it is which hormones our bodies produce that determines sex, without regard to whether our bodies produce them with or without deliberate intervention. I don't particularly care if we settle on one or the other but it seems somewhat disingenuous to imply that this has never mattered when we've been running with a definition of sex that assumed a natural expression of hormones.
The significant point in many of these discussions is glossing over the outliers, the 2% whose hardware+wetware doesn't line up with the majority of humans.
Historically, globally, throughout history there has always been such outliers (Gender in Bugis-Makassar society, Sistagirls in Torres Strait groups, etc).
> when we've been running with a definition of
The "we" in a global setting and sans context is an issue, there's a plurality of definitions that are used depending on focus and degree of technical detail.
Richard Dawkins use of "sex" in recent interviews is a classic example, he deftly occludes the 2% of humans that don't neatly fall into M|F buckets by taking a strictly reproductive evolutionary razor to the question and responds (as best I recall) that "there are only two (reproductive) sexes, Males (viable sperm producers), and Females (viable egg producers) .. that's just (reproductive) biology".
That's a bit of a mean spirited one way mapping from two reproductive buckets to those humans that cleanly fall into those buckets.
Dawkins avoids and hopes the audience doesn't press on whether all humans fit into one or the other bucket (they don't) and of course his interviewer at the time doesn't press.
To give you an example of what sex means: The level of testosterone in my body may be influenced by society, but if you measure it, you will have no doubt telling my sex. Like for most adults, it sets me apart by an order of magnitude from the other sex. And I make this prediction without ever having measured. Any biologist would validate it just from looking at me.
It is very likely that my level of testosterone will never be in the range of the other sex. Since puberty, until death. Society cannot change that. The diurnal cycle? You jest.
> Until recently, a model like the one Menon’s team employed would help researchers sort brains into different groups but wouldn’t provide information about how the sorting happened. Today, however, researchers have access to a tool called “explainable AI,” which can sift through vast amounts of data to explain how a model’s decisions are made.
> Using explainable AI, Menon and his team identified the brain networks that were most important to the model’s judgment of whether a brain scan came from a man or a woman. They found the model was most often looking to the default mode network, striatum, and the limbic network to make the call.