This article reads like a longform version of why statistical power/effect size is important but doesn't mention it by name, only alluding to it. I suspect there would be much less of a replication crisis if, instead of just focusing on statistical significance, there was also a focus on effect size, or both.
My impression is that Nautilus and Quanta are, among pop science magazines, fairly high brow, but that Nautilus is substantially more likely to use human-interest narratives (e.g., the opening vignette) to keep the reader's attention. Is that right, or do I have the wrong impression?
You'll never get a full brain scan with fMRIs. For real detail, you'll need a planetary-scale XFEL with a human head-sized reticule and be okay with your irl head being vaporized.
Optimized models for simulating your brain accurately don't exist yet either. I highly suspect that we will have to finish the above first.
Longer version. It's still a copy, but is a copy of you still you; are you still you if you've had a copy? (My belief is yes; though at that point timelines diverge and neither 'you' is fungible for the other.)
As a cognitive scientist with only 1 graduate level fMRI course and no hands-on experience outside of playing around with Neurodebian, I will not be so bold (haha) as to make a claim that there is absolutely nothing to fMRI.
However, I am very skeptical.
I would not be willing to make any strong claims about causality when the alleged neural event and the measurement thereof are 15-30 seconds apart, and the thing being measured (blood flow) is not the event itself (neural activity), and there are so, so many potential sources of error.
Now, structural MRI is super cool.
I have a bunch of friends at my university who work in fMRI brain imaging. I feel bad, but when we talk about their work, I can't help but think they are chasing alchemy. Or maybe a better analogy is that it is almost like they are trying to learn about the world (brain, cognition) based on the shadows at the back of Plato's cave (fMRI BOLD signals).
I sincerely hope they overcome the challenges, but I expect they will be facing the exact or nearly the same challenges 20 years from now.
Even if they’re on a deadend they may discover a thing or two that can be applied somewhere else. But I hear you, your intuition is probably right. But if they left this area unexplored it would still be a loss.
It's a while since I ran fMRI studies but my favourite analogy was to tell people that it was like trying to measure your electricity usage by watching your water meter.
Tom Mitchell has been using machine learning to interpret fMRI data.[1] The method struck me as one that could potentially work (though I was a bit surprised by the apparent success rate of studies in his lab).
The fundamental problem with fMRI is its poor time resolution. MEG is promising in this respect.[2]
> "Functional magnetic resonance imaging has transformed medicine."
This is not even close to true. The clinical impact of fMRI is very constrained. I've met many clinicians who have completely written it off at this point. Others that remain interested but find it mostly impractical and error prone. The only people I've met who seem to have any significant investment in it as a technology are cognitive scientists, and that's a ways from medicine.
> It allows non-invasive mapping of a patient’s brain regions to enable more accurate, precise neurosurgery,1"
Approximately nobody does this. fMRI isn't really even on the radar for most neurosurgeons. It's only recently that even detailed structural information (e.g. DTI) has got some real traction, let alone functional info. Note commercial vendors have had clinical packages on the scanners for (nearly? I forget who shipped DWI stuff first) a couple decades now. No such thing exists for fMRI, at minimum you need to buy a 3rd party processing system and may need a research key on your scanner.
Some researchers are interested in this; that's a long way from transforming anything.
> " as well as validating pharmacological effects of potential drugs on human brains.2"
There is perhaps a bit more promise here, but it's still way into the research only end. In the 15 years since the referenced paper was published, not much of this "promise" has been realized as far as I know (but this is further from my wheelhouse)
I was just reading about them using fMRI in conjunction with high frequency focused ultrasound to lesion parts of the brain associated with certain mental disorders such as OCD. They seemed to be having decent success with that
Sure, there is research all over the place - but HIFU also hasn’t made a real impact clinically, maybe it will in future.
Most of this stuff gets interest in papers for a few years , turns out to be impractical for whatever reason and dies on the vine. That’s ok, it’s worth trying .
People doing related research are interested in it, and a few teaching hospitals etc., but that’s a really small fraction. it’s not really a factor in most clinical practice.
Thus “transformed medicine” is more than a bit of a stretch.
> it’s not really a factor in most clinical practice.
For a nice anecdata point - I’m an MR technologist and clinically I’ve seen it done once, I’ve done it once and I’ve heard a colleague mention doing one.
I’ve done them as research scans a hell of a lot of times.
fMRI is used to plan procedures that put eloquent cortex at risk. For example, tumor resections. Typically, fMRI is used for initial planning (whether the tumor is in the dominant hemisphere, how to approach the tumor). Then, in the actual surgery, function is confirmed using awake cortical mapping (e.g. stimulating different cortical areas with a bipolar electrode while the patient performs language tasks).
No, it is not. In my experience it is at best rarely used for planning, i.e. typically planning is done without fMRI. It will probably have DWI but maybe not tensor information (although that is growing).
Awake cortical mapping and/or task based references are very typical (near 1/2 of procedures, iirc). fMRI is not.
I'm not saying it's a bad idea or anything. I have a pretty good understanding of all the tradeoffs for fMRI but I'm not arguing for or against it. I'm reacting to the claim that is commonly deployed in clinical (not research) practice, which just doesn't match my (fairly extensive) experience.
So the articles claims of "transforming medicine" are aspirational.
source: Have worked directly in this market (clinical), hashed this out with many neurosurgeons,seen the procedures, etc. See also MR tech comment in this thread, in my experience very typical.
I don't really know what you are responding to here.
As I posited, there is research interest and a few people doing it clinically. Your links support this. What they don't support is the idea that this is typical clinical practice.
For those interested, here's a nice Tweet and video from a neurosurgeon at the Mayo clinic in Arizona demonstrating the use of both fMRI and intraoperative motor mapping: "Preoperative and Intraoperative mapping help with selection, safety and strategy."
That’s interesting! You are right the resolution (both spatial and temporal) and accurately coregistering (and then with surgery, brain shift) can be among the limiting factors for clinical use. Plus the difficult in getting good signal.
I suspect this has more to do with where they had it done than anything else. There are neurosurgeons who champion it for sure , but not that many. Most places aren’t set up for it at all from what I’ve seen. I'm guessing her procedure wasnt' done awake, was it? This is a very common way to deal with the functional stuff especially near speech or motor.
Awake cortical mapping has much better resolution than fMRI, and it avoids image registration issues, but it requires a great deal of planning and patient motivation. If it turns out that the language center is on the opposite side from the area of the surgery, then awake mapping might not even be necessary. This is why fMRI is often used as a planning step before awake cortical mapping during the actual surgery.
The other thing to consider here is unintentional media bias - when a lay publication wants information they tend to go to tier 1 research & academic institutions; exactly the sort of place that has all the latest new toys. This is often very far from typical practice. So people outside of healthcare hear about all sort of things that are a decade or more from general usage, even if they get there eventually.
From the outside, it's probably difficult to understand how conservative medicine is generally. Even effective new practices often take decades to become really mainstream.
I'm not on the outside of medicine. Significant practice variation exists, but complex neurosurgical conditions are usually managed at academic centers and fMRI is used for planning in neurooncology [1] and epilepsy surgery [2].
Fair enough, I didn't mean to suggest you were I was speaking more generally but worded that poorly.
My comments come from systems supporting thousands of clinical neurooncological procedures (i.e. tumor resections) in planning and execution with very little interest or utilization of fMRI beyond a handful proponents and their sites. Quite literally barely on the radar of most of the neurosurgeons apart from occasional papers, and some of them are quite negative about it also.
I could have an inaccurate picture of the breadth of clinical practice, and it's certainly a couple years out of date, but I would be very surprised to find a huge upsurge of usage outside research had happened.
It is certainly the case that the articles claim of "transformation" hasn't happened in that space.
If the article had instead claimed that some of the trickiest cases tend to have fMRI done (true, surgeons will take all they help they can get trying cases that otherwise might be inoperable) or that they are a feature of high profile academic sites (also mostly true) I wouldn't have objected.
I respect your experience, and I'll concede that fMRI for pre-surgical planning in neurooncology is used at only some centers. If you worked mainly in neurooncology you may have missed some of the uses of fMRI for epilepsy surgery. Thanks for the discussion.
> [...] exactly the sort of place that has all the latest new toys. This is often very far from typical practice. So people outside of healthcare hear about all sort of things that are a decade or more from general usage, even if they get there eventually.
To give a bit more context to my anecdata.
The surgery I was talking about was performed 10 years ago. In Poland. In city with around 700k people. In probably the best hospital for such surgeries in this city with only one other hospital also doing neurosurgery in this city but specializing more in the spinal surgery. The neurosurgical ward back then was in dire need of renovations so not exactly shiny new place that has funds for the best toys.
As I said it looked pretty routine back then.
She did not have fMRI for her following surgeries because tumors were not close to the speech center (opposite hemisphere, then frontal lobes).
Interesting; I've much more exposure to US. Overall, I'd say there are sites that do it pretty routinely, but far more sites that basically never do it. And others that don't even have the equipment/licenses if they wanted to. For what it's worth, overall neurosurgeons do much more spine surgery than brain. In the particular case you mention, I've seen a lot of mapping and/or testing for surgeries near eloquent function, but little fMRI. Outside of that, basically unheard of.
> This is why fMRI is often used as a planning step before awake cortical mapping during the actual surgery.
That must have been the case.
I guess surgeons were trying to do whatever was in their power to not damage her speech center. Right after operation she had a few brief seizures when she lost ability to speak for a minute or two, but as the damage healed they quickly stopped.
Thank you for sharing your story. Although fMRI has low resolution, it can be very useful for surgical planning in certain clinical scenarios. For example, to determine hemispheric dominance: "Although the estimated percentages are of some debate, language is the purview of the left hemisphere in approximately 95% of right-handed people and 70% of left-handed people ... At MSKCC, language lateralization mapping is most often requested in right-handed patients with left hemispheric lesions, left-handed patients with left or right hemispheric lesions, or right-handed patients with right hemispheric lesions and signs or symptoms of aphasia." [1]
Oh that's a good point, I should have been more clear (and honestly had forgotten BrainWave, you can option that with their DTI fibre stuff).
These systems are still not on the scanner (unless some of the latest acquisition stations support it? Still not on most of the deployed stuff) so typically another workstation is needed at least, this package does happens to be from the vendor. For anyone reading who finds this confusing, you can license on the machine the ability to do the "pulse sequences" you need, but it only handles the raw data - you typically need another step to process that into images you can use.
Otoh most of the people I know doing this on GE scanners used something else, but that could easily be sampling bias.
> As for the Eklund, et al. paper, Rosen conceded, “It was an excellent point of statistics. It’s actually a point we understood, but there are no doubt lots of people that didn’t understand.” But he thinks the paper’s implications were overblown. “The impact of that paper was actually pretty modest in terms of the number of results that were invalidated, that were important results. And whether we were significantly misleading people or sending doctors astray or anything like that, was negligible as best as I could tell. But the paper got a lot of press, and suddenly, now fMRI has a black eye.”
“You see, it’s the others that don’t do it right, we do things correctly so _our_ results are accurate and relevant. It’s those other researchers who are not contributing to curing cancer/Alzheimer’s/etc” - this is the argument I have heard time and again from many colleagues when we point some fundamental issue with their whole field. Rest assured they also do the same mistakes but have often built large cognitive fortresses to shield their conscience from the reality that they are also a charlatan at least partially.
The only endgame that you can truly live with given such realisations is to leave the field but many can’t afford to do that, so you have a perpetuation of such fields which shouldn’t exist at the scale that they do.
1. fMRI only measure brain activity indirectly.
2. Nothing stops people from messing with any one parameter, setting, statistical knob out of houndreds.
I don't know if #1 is a problem itself. We measure many things indirectly. As long as the limitations are known and respected, it is still useful.
#2 Is the real issue. What we need is a movement towards saving and publishing the the actual data as close to the hardware as possible. But I'm sure medical hardware is full of quirky, proprietary Windows CE junk that makes this almost impossible.
I don't know what the solution is. If the healthcare industry wanted better software they would have to pay even more than they already do. Cheaper and better software doesn't happen because those companies and developers would rather work elsewhere. Real disruptors will have a hard time convincing the 100 vendors they need to integrate with to change their ways.
It seems that large industries will always be 10-15 years behind 'what they should be'.
What we need is a movement towards saving and publishing the the actual data as close to the hardware as possible. But I'm sure medical hardware is full of quirky, proprietary Windows CE junk that makes this almost impossible.
It's always easy to blame proprietary junk. But in this case, from the couple of things I know about it, I don't think the real problem in publishing actual data is that it's proprietary. Rather that doing something with that data is very hard. To the point that if you don't know how to build an MRI scanner, you probably also don't know how to transform the raw data into a usable and correct image.
Its an intetesting article from a personal prespective. I found it very frustrating all the imagery was essentially the same graphic presented multiple times. Show a real brain scan! or a picture of thr bloody machine.
52 comments
[ 1.6 ms ] story [ 104 ms ] threadOptimized models for simulating your brain accurately don't exist yet either. I highly suspect that we will have to finish the above first.
Longer version. It's still a copy, but is a copy of you still you; are you still you if you've had a copy? (My belief is yes; though at that point timelines diverge and neither 'you' is fungible for the other.)
However, I am very skeptical.
I would not be willing to make any strong claims about causality when the alleged neural event and the measurement thereof are 15-30 seconds apart, and the thing being measured (blood flow) is not the event itself (neural activity), and there are so, so many potential sources of error.
Now, structural MRI is super cool.
I have a bunch of friends at my university who work in fMRI brain imaging. I feel bad, but when we talk about their work, I can't help but think they are chasing alchemy. Or maybe a better analogy is that it is almost like they are trying to learn about the world (brain, cognition) based on the shadows at the back of Plato's cave (fMRI BOLD signals).
I sincerely hope they overcome the challenges, but I expect they will be facing the exact or nearly the same challenges 20 years from now.
As a fellow cognitive scientist, I am too.
The fundamental problem with fMRI is its poor time resolution. MEG is promising in this respect.[2]
[1] https://www.youtube.com/watch?v=zy0-_HNF1us&t=6m8s
[2] https://en.wikipedia.org/wiki/Magnetoencephalography
This is not even close to true. The clinical impact of fMRI is very constrained. I've met many clinicians who have completely written it off at this point. Others that remain interested but find it mostly impractical and error prone. The only people I've met who seem to have any significant investment in it as a technology are cognitive scientists, and that's a ways from medicine.
> It allows non-invasive mapping of a patient’s brain regions to enable more accurate, precise neurosurgery,1"
Approximately nobody does this. fMRI isn't really even on the radar for most neurosurgeons. It's only recently that even detailed structural information (e.g. DTI) has got some real traction, let alone functional info. Note commercial vendors have had clinical packages on the scanners for (nearly? I forget who shipped DWI stuff first) a couple decades now. No such thing exists for fMRI, at minimum you need to buy a 3rd party processing system and may need a research key on your scanner.
Some researchers are interested in this; that's a long way from transforming anything.
> " as well as validating pharmacological effects of potential drugs on human brains.2"
There is perhaps a bit more promise here, but it's still way into the research only end. In the 15 years since the referenced paper was published, not much of this "promise" has been realized as far as I know (but this is further from my wheelhouse)
Most of this stuff gets interest in papers for a few years , turns out to be impractical for whatever reason and dies on the vine. That’s ok, it’s worth trying .
Except at basically any academic neurosurgery center?
https://thejns.org/focus/view/journals/neurosurg-focus/48/2/...
People doing related research are interested in it, and a few teaching hospitals etc., but that’s a really small fraction. it’s not really a factor in most clinical practice.
Thus “transformed medicine” is more than a bit of a stretch.
For a nice anecdata point - I’m an MR technologist and clinically I’ve seen it done once, I’ve done it once and I’ve heard a colleague mention doing one. I’ve done them as research scans a hell of a lot of times.
See more: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4966674/
No, it is not. In my experience it is at best rarely used for planning, i.e. typically planning is done without fMRI. It will probably have DWI but maybe not tensor information (although that is growing).
Awake cortical mapping and/or task based references are very typical (near 1/2 of procedures, iirc). fMRI is not.
I'm not saying it's a bad idea or anything. I have a pretty good understanding of all the tradeoffs for fMRI but I'm not arguing for or against it. I'm reacting to the claim that is commonly deployed in clinical (not research) practice, which just doesn't match my (fairly extensive) experience.
So the articles claims of "transforming medicine" are aspirational.
source: Have worked directly in this market (clinical), hashed this out with many neurosurgeons,seen the procedures, etc. See also MR tech comment in this thread, in my experience very typical.
[1] https://thejns.org/focus/view/journals/neurosurg-focus/48/2/...
[2] https://n.neurology.org/content/88/4/395.long
As I posited, there is research interest and a few people doing it clinically. Your links support this. What they don't support is the idea that this is typical clinical practice.
https://twitter.com/BernardBendokMD/status/13785056756894556...
It wasn't a big deal, so I think they apply it routinely in some cases in their practice.
I'm not sure if they got much from it because it was really low res in comparison with standard MRI.
Although she made a full recovery after that surgery.
I suspect this has more to do with where they had it done than anything else. There are neurosurgeons who champion it for sure , but not that many. Most places aren’t set up for it at all from what I’ve seen. I'm guessing her procedure wasnt' done awake, was it? This is a very common way to deal with the functional stuff especially near speech or motor.
Glad to hear she had a good outcome!
It really isn't - see my other comment.
The other thing to consider here is unintentional media bias - when a lay publication wants information they tend to go to tier 1 research & academic institutions; exactly the sort of place that has all the latest new toys. This is often very far from typical practice. So people outside of healthcare hear about all sort of things that are a decade or more from general usage, even if they get there eventually.
From the outside, it's probably difficult to understand how conservative medicine is generally. Even effective new practices often take decades to become really mainstream.
[1] https://thejns.org/focus/view/journals/neurosurg-focus/48/2/...
[2] https://n.neurology.org/content/88/4/395.long
Fair enough, I didn't mean to suggest you were I was speaking more generally but worded that poorly.
My comments come from systems supporting thousands of clinical neurooncological procedures (i.e. tumor resections) in planning and execution with very little interest or utilization of fMRI beyond a handful proponents and their sites. Quite literally barely on the radar of most of the neurosurgeons apart from occasional papers, and some of them are quite negative about it also.
I could have an inaccurate picture of the breadth of clinical practice, and it's certainly a couple years out of date, but I would be very surprised to find a huge upsurge of usage outside research had happened.
It is certainly the case that the articles claim of "transformation" hasn't happened in that space.
If the article had instead claimed that some of the trickiest cases tend to have fMRI done (true, surgeons will take all they help they can get trying cases that otherwise might be inoperable) or that they are a feature of high profile academic sites (also mostly true) I wouldn't have objected.
For me fMRI falls into the cool-but-has-some-issues (mainly around consistency and processing) side of tech.
And yes, thanks for the discussion.
> [...] exactly the sort of place that has all the latest new toys. This is often very far from typical practice. So people outside of healthcare hear about all sort of things that are a decade or more from general usage, even if they get there eventually.
To give a bit more context to my anecdata.
The surgery I was talking about was performed 10 years ago. In Poland. In city with around 700k people. In probably the best hospital for such surgeries in this city with only one other hospital also doing neurosurgery in this city but specializing more in the spinal surgery. The neurosurgical ward back then was in dire need of renovations so not exactly shiny new place that has funds for the best toys.
As I said it looked pretty routine back then.
She did not have fMRI for her following surgeries because tumors were not close to the speech center (opposite hemisphere, then frontal lobes).
Interesting; I've much more exposure to US. Overall, I'd say there are sites that do it pretty routinely, but far more sites that basically never do it. And others that don't even have the equipment/licenses if they wanted to. For what it's worth, overall neurosurgeons do much more spine surgery than brain. In the particular case you mention, I've seen a lot of mapping and/or testing for surgeries near eloquent function, but little fMRI. Outside of that, basically unheard of.
She had that too.
> This is why fMRI is often used as a planning step before awake cortical mapping during the actual surgery.
That must have been the case.
I guess surgeons were trying to do whatever was in their power to not damage her speech center. Right after operation she had a few brief seizures when she lost ability to speak for a minute or two, but as the damage healed they quickly stopped.
[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4966674/
That was the case for my SO.
This is not the case, for example see the GE BrainWave software.
https://www.gehealthcare.com/products/advanced-visualization...
These systems are still not on the scanner (unless some of the latest acquisition stations support it? Still not on most of the deployed stuff) so typically another workstation is needed at least, this package does happens to be from the vendor. For anyone reading who finds this confusing, you can license on the machine the ability to do the "pulse sequences" you need, but it only handles the raw data - you typically need another step to process that into images you can use.
Otoh most of the people I know doing this on GE scanners used something else, but that could easily be sampling bias.
> As for the Eklund, et al. paper, Rosen conceded, “It was an excellent point of statistics. It’s actually a point we understood, but there are no doubt lots of people that didn’t understand.” But he thinks the paper’s implications were overblown. “The impact of that paper was actually pretty modest in terms of the number of results that were invalidated, that were important results. And whether we were significantly misleading people or sending doctors astray or anything like that, was negligible as best as I could tell. But the paper got a lot of press, and suddenly, now fMRI has a black eye.”
“You see, it’s the others that don’t do it right, we do things correctly so _our_ results are accurate and relevant. It’s those other researchers who are not contributing to curing cancer/Alzheimer’s/etc” - this is the argument I have heard time and again from many colleagues when we point some fundamental issue with their whole field. Rest assured they also do the same mistakes but have often built large cognitive fortresses to shield their conscience from the reality that they are also a charlatan at least partially.
The only endgame that you can truly live with given such realisations is to leave the field but many can’t afford to do that, so you have a perpetuation of such fields which shouldn’t exist at the scale that they do.
1. fMRI only measure brain activity indirectly. 2. Nothing stops people from messing with any one parameter, setting, statistical knob out of houndreds.
I don't know if #1 is a problem itself. We measure many things indirectly. As long as the limitations are known and respected, it is still useful.
#2 Is the real issue. What we need is a movement towards saving and publishing the the actual data as close to the hardware as possible. But I'm sure medical hardware is full of quirky, proprietary Windows CE junk that makes this almost impossible.
I don't know what the solution is. If the healthcare industry wanted better software they would have to pay even more than they already do. Cheaper and better software doesn't happen because those companies and developers would rather work elsewhere. Real disruptors will have a hard time convincing the 100 vendors they need to integrate with to change their ways.
It seems that large industries will always be 10-15 years behind 'what they should be'.
It's always easy to blame proprietary junk. But in this case, from the couple of things I know about it, I don't think the real problem in publishing actual data is that it's proprietary. Rather that doing something with that data is very hard. To the point that if you don't know how to build an MRI scanner, you probably also don't know how to transform the raw data into a usable and correct image.
Source: I know a guy who worked on Alzheimer's and they combined several radiology sources to form a composite brain scan to study structure.