People used to worry that robots were getting so smart that they’d soon start secretly plotting to take over the world. But now experts worry that AI is getting so smart that it could be secretly plotting to do racism to Black people:
However, our findings that AI can trivially predict self-reported race — even from corrupted, cropped, and noised medical images — in a setting where clinical experts cannot, creates an enormous risk for all model deployments in medical imaging: if an AI model secretly used its knowledge of self-reported race to misclassify all Black patients, radiologists would not be able to tell using the same data the model has access to.
From a new preprint on arXiv:
Imon Banerjee, Ananth Reddy Bhimireddy, John L. Burns, Leo Anthony Celi, Li-Ching Chen, Ramon Correa, Natalie Dullerud, Marzyeh Ghassemi, Shih-Cheng Huang, Po-Chih Kuo, Matthew P Lungren, Lyle Palmer, Brandon J Price, Saptarshi Purkayastha, Ayis Pyrros, Luke Oakden-Rayner, Chima Okechukwu, Laleh Seyyed-Kalantari, Hari Trivedi, Ryan Wang, Zachary Zaiman, Haoran Zhang, Judy W Gichoya
Background: In medical imaging, prior studies have demonstrated disparate AI performance by race, yet there is no known correlation for race on medical imaging that would be obvious to the human expert interpreting the images.
Methods: Using private and public datasets we evaluate: A) performance quantification of deep learning models to detect race from medical images, including the ability of these models to generalize to external environments and across multiple imaging modalities, B) assessment of possible confounding anatomic and phenotype population features, such as disease distribution and body habitus as predictors of race, and C) investigation into the underlying mechanism by which AI models can recognize race.
Findings: Standard deep learning models can be trained to predict race from medical images with high performance across multiple imaging modalities. Our findings hold under external validation conditions, as well as when models are optimized to perform clinically motivated tasks. We demonstrate this detection is not due to trivial proxies or imaging-related surrogate covariates for race, such as underlying disease distribution. Finally, we show that performance persists over all anatomical regions and frequency spectrum of the images suggesting that mitigation efforts will be challenging and demand further study.
Interpretation: We emphasize that model ability to predict self-reported race is itself not the issue of importance. However, our findings that AI can trivially predict self-reported race — even from corrupted, cropped, and noised medical images — in a setting where clinical experts cannot, creates an enormous risk for all model deployments in medical imaging: if an AI model secretly used its knowledge of self-reported race to misclassify all Black patients, radiologists would not be able to tell using the same data the model has access to.
From the blog of one of the authors:
AUGUST 2, 2021 ~ LUKEOAKDENRAYNER
Is this the darkest timeline? Are we the baddies?
… instead I wanted to write something else which I think will complement the paper; an explanation of why I and many of my co-authors think this issue is important.
One thing we noticed when we were working on this research was that there was a clear divide in our team. The more clinical and safety/bias related researchers were shocked, confused, and frankly horrified by the results we were getting. Some of the computer scientists and the more junior researchers on the other hand were surprised by our reaction. They didn’t really understand why we were concerned.
So in a way, this blog post can be considered a primer, a companion piece for the paper which explains the why. Sure, AI can detect a patient’s racial identity, but why does it matter?
Disclaimer: I’m white. I’m glad I got to contribute, and I am happy to write about this topic, but that does not mean I am somehow an authority on the lived experiences of minoritized racial groups. These are my opinions after discussion with my much more knowledgeable colleagues, several of whom have reviewed the blog post itself.
A brief summary
In extremely brief form, here is what the paper showed:
- AI can trivially learn to identify the self-reported racial identity of patients to an absurdly high degree of accuracy
- AI does learn to do this when trained for clinical tasks
- These results generalise, with successful external validation and replication in multiple x-ray and CT datasets
- Despite many attempts, we couldn’t work out what it learns or how it does it. It didn’t seem to rely on obvious confounders, nor did it rely on a limited anatomical region or portion of the image spectrum.
Now for the important part: so what?
An argument in four steps
I’m going to try to lay out, as clearly as possible, that this AI behaviour is both surprising, and a very bad thing if we care about patient safety, equity, and generalisability.
The argument will have the following parts:
- Medical practice is biased in favour of the privileged classes in any society, and worldwide towards a specific type of white men.
- AI can trivially learn to recognise features in medical imaging studies that are strongly correlated with racial identity. This provides a powerful and direct mechanism for models to incorporate the biases in medical practice into their decisions.
- Humans cannot identify the racial identity of a patient from medical images. In medical imaging we don’t routinely have access to racial identity information, so human oversight of this problem is extremely limited at the clinical level.
- The features the AI makes use of appear to occur across the entire image spectrum and are not regionally localised, which will severely limit our ability to stop AI systems from doing this.
There are several other things I should point out before we get stuck in. First of all, a definition. We are talking about racial identity, not genetic ancestry or any other biological process that might come to mind when you hear the word “race”. Racial identity is a social, legal, and political construct that consists of our own perceptions of our race, and how other people see us. In the context of this work, we rely on self-reported race as our indicator of racial identity.
Before you jump in with questions about this approach and the definition, a quick reminder on what we are trying to research. Bias in medical practice is almost never about genetics or biology. No patient has genetic ancestry testing as part of their emergency department workup. We are interested in factors that may bias doctors in how they decide to investigate and treat patients, and in that setting the only information they get is visual (i.e., skin tone, facial features etc.) and sociocultural (clothing, accent and language use, and so on). What we care about is race as a social construct, even if some elements of that construct (such as skin tone) have a biological basis.
Secondly, whenever I am using the term bias in this piece, I am referring to the social definition, which is a subset of the strict technical definition; it is the biases that impact decisions made about humans on the basis of their race. These biases can in turn produce health disparities, which the NIH defines as “a health difference that adversely affects disadvantaged populations“.
Third, I want to take as given that racial bias in medical AI is bad. I feel like this shouldn’t need to be said, but the ability of AI to homogenise, institutionalise, and algorithm-wash health disparities across regions and populations is not a neutral thing.
AI can seriously make things much, much worse.
… In medical imaging we like to think of ourselves as above this problem, particularly with respect to race because we usually don’t know the identity of our patients. We report the scans without ever seeing the person, but that only protects us from direct bias. Biases still affect who gets referred for scans and who doesn’t, and they affect which scans are ordered. …
But it is true that, in general, we read the scan as it comes. The scan can’t tell us what colour a person’s skin is.
Part II – AI can detect racial identity in x-rays and CT scans
I’ve already included some results up in the summary section, and there are more in the paper, but I’ll very briefly touch on my interpretation of them here.
Firstly, the performance of these models ranges from high to absurd. An AUC of 0.99 for recognising the self-reported race of a patient, which has no recognised medical imaging correlate? This is flat out nonsense.
Every radiologist I have told about these results is absolutely flabbergasted, because despite all of our expertise, none of us would have believed in a million years that x-rays and CT scans contain such strong information about racial identity. Honestly we are talking jaws dropped – we see these scans everyday and we have never noticed.
The second important aspect though is that, with such a strong correlation, it appears that AI models learn the features correlated with racial identity by default. For example, in our experiments we showed that the distribution of diseases in the population for several datasets was essentially non-predictive of racial identity (AUC = 0.5 to 0.6), but we also found that if you train a model to detect those diseases, the model learns to identify patient race almost as well as the models directly optimised for that purpose (AUC = 0.86). Whaaat?
Despite racial identity not being useful for the task (since the disease distribution does not differentiate racial groups), the model learns it anyway? …
But no matter how it works, the take-home message is that it appears that models will tend to learn to recognise race, even when it seems irrelevant to the task. So the dozens upon dozens of FDA approved x-ray and CT scan AI models on the market now … probably do this^^? Yikes!
There is one more interpretation of these results that is worth mentioning, for the “but this is expected model behaviour” folks. Even from a purely technical perspective, ignoring the racial bias aspect, the fact models learn features of racial identity is bad. There is no causal pathway linking racial identity and the appearance of, for example, pneumonia on a chest x-ray. By definition these features are spurious.
They are shortcuts. Unintended cues. The model is underspecified for the problem it is intended to solve.
However we want to frame this, the model has learned something that is wrong, and this means the model can behave in undesirable and unexpected ways.
I won’t be surprised if this becomes a canonical example of the biggest weakness of deep learning – the ability of deep learning to pick up unintended cues from the data. I’m certainly going to include it in all my talks.
Part III – Humans can’t identify racial identity in medical images
… The problem is much worse for racial bias. At least in MRI super-resolution, the radiologist is expected to review the original low quality image to ensure it is diagnostic quality (which seems like a contradiction to me, but whatever). In AI with racial bias though, humans literally cannot recognise racial identity from images^^^. Unless they are provided with access to additional data (which they don’t currently have easy access to in imaging workflows) they will be completely unable to appreciate the bias no matter how skilled they are and no matter how much effort they apply to the task.
Part IV – We don’t know how to stop it
This is probably the biggest problem here. We ran an extensive series of experiments to try and work out what was going on.
First, we tried obvious demographic confounders (for example, Black patients tend to have higher rates of obesity than white patients, so we checked whether the models were simply using body mass/shape as a proxy for racial identity). None of them appeared to be responsible, with very low predictive performance when tested alone.
Next we tried to pin down what sort of features were being used. There was no clear anatomical localisation, no specific region of the images that contributed to the predictions. Even more interesting, no part of the image spectrum was primarily responsible either. We could get rid of all the high-frequency information, and the AI could still recognise race in fairly blurry (non-diagnostic) images. Similarly, and I think this might be the most amazing figure I have ever seen, we could get rid of the low-frequency information to the point that a human can’t even tell the image is still an x-ray, and the model can still predict racial identity just as well as with the original image!
Damn their eyes!
Performance is maintained with the low pass filter to around the LPF25 level, which is quite blurry but still readable. But for the high-pass filter, the model can still recognise the racial identity of the patient well past the point that the image is just a grey box …
This difficulty in isolating the features associated with racial identity is really important, because one suggestion people tend to have when they get shown evidence of racial bias is that we should make the algorithms “colorblind” – to remove the features that encode the protected attribute and thereby make it so the AI cannot “see” race but should still perform well on the clinical tasks we care about.
Here, it seems like there is no easy way to remove racial information from images. It is everywhere and it is in everything.
Perhaps Disraeli was right when he had the character who was his mouthpiece in his novels explain, “All is race.”
An urgent problem
AI seems to easily learn racial identity information from medical images, even when the task seems unrelated. We can’t isolate how it does this, and we humans can’t recognise when AI is doing it unless we collect demographic information (which is rarely readily available to clinical radiologists). That is bad.
There are around 30 AI systems using CXR and CT Chest imaging on the market currently, FDA cleared, many of which were trained on the exact same datasets we utilised in this research. That is worse.
I don’t know about you, but I’m worried. AI might be superhuman, but not every superpower is a force for good.
The line between superheroism and supervillainy is a fine one.
It’s almost as if race does exist. But of course we’ve been told over and over that that can’t possibly be true. But did anybody tell Artificial Intelligence that? It’s almost as if AI isn’t a True Believer in the conventional wisdom about the scientific nonexistence of race. Something must be done to inject the natural stupidity of our elite wisdom into Artificial Intelligence.