Machine learning classification of functional neurological disorder using structural brain MRI features 2024 Westlin, Perez et al

Andy

Andy

Retired committee member
Abstract

Background
Brain imaging studies investigating grey matter in functional neurological disorder (FND) have used univariate approaches to report group-level differences compared with healthy controls (HCs). However, these findings have limited translatability because they do not differentiate patients from controls at the individual-level.

Methods
183 participants were prospectively recruited across three groups: 61 patients with mixed FND (FND-mixed), 61 age-matched and sex-matched HCs and 61 age, sex, depression and anxiety-matched psychiatric controls (PCs). Radial basis function support vector machine classifiers with cross-validation were used to distinguish individuals with FND from HCs and PCs using 134 FreeSurfer-derived grey matter MRI features.

Results
Patients with FND-mixed were differentiated from HCs with an accuracy of 0.66 (p=0.005; area under the receiving operating characteristic (AUROC)=0.74); this sample was also distinguished from PCs with an accuracy of 0.60 (p=0.038; AUROC=0.56). When focusing on the functional motor disorder subtype (FND-motor, n=46), a classifier robustly differentiated these patients from HCs (accuracy=0.72; p=0.002; AUROC=0.80). FND-motor could not be distinguished from PCs, and the functional seizures subtype (n=23) could not be classified against either control group. Important regions contributing to statistically significant multivariate classifications included the cingulate gyrus, hippocampal subfields and amygdalar nuclei. Correctly versus incorrectly classified participants did not differ across a range of tested psychometric variables.

Conclusions
These findings underscore the interconnection of brain structure and function in the pathophysiology of FND and demonstrate the feasibility of using structural MRI to classify the disorder. Out-of-sample replication and larger-scale classifier efforts incorporating psychiatric and neurological controls are needed.

Open access, https://jnnp.bmj.com/content/early/2024/07/20/jnnp-2024-333499
 
Perez is just as awful and biased, but the data about differentiating are pretty bad. Not sure if this is anything beyond "there are non-specific changes", which is not super useful but does invalidate the entire premise of the ideology.

But a faith-based ideology doesn't adapt to facts, so it's not as it matters much. We are talking about people who have no issues with insisting that all the symptoms are caused by deconditioning, but will excuse the failure of an exercise program on the basis that they are active enough anyway. They don't care about facts, and neither does the rest of the profession, happy to let them spew bullshit as if it somehow frees the entire profession from responsibility here.
 
Abstract said:
134 FreeSurfer-derived grey matter MRI features.
Click to expand...
So many options for correlations, and even then, there wasn't that much of note. I'm pretty sure that if you applied normal distribution probabilities for 134 features to 180 individuals randomly put into 3 groups, most of the time you would find some combination of some features that separated the groups at least as well as they found here.

Accuracy
FND mixed vs HC 0.66
FND mixed vs PC 0.60
FND-motor vs HC 0.72
FND-motor vs PC not distinguishable
FND-seizure vs HC not distinguishable
FND-seizure vs PC not distinguishable
Having particular psychometric characteristics made no difference to the likelihood of the model placing the person in the right category


Just maybe the finding that "functional" motor disorder can be identified by structural brain features visible in MRI will be replicated, but with that many MRI features, the identified relationships are probably just noise. Despite their results suggesting one type of FND might be correlated with brain structure characteristics observable on an MRI and one type is not, from the abstract the authors strangely still seem set on suggesting that both "conditions" are the same thing (FND), presumably with the same etiology, and presumably with the same treatment approach. It makes no sense.

The presence or absence of something structural that can be seen on a standard MRI of course doesn't prove or disprove that the condition is psychosomatic - things like migraine and epilepsy don't show up, so it's not at all surprising that "functional" seizures don't either.
 
But I still don't get how they can argue there could be structure-related biomarkers and still argue that it's all a software issue. Why don't they address that contradiction? At some point, don't they need to acknowledge that something is physically wrong besides "software" and "brain network" issues?
 
They can still argue that it is something about a patient's thoughts and/or the behaviour that flows from that faulty thinking that is causing both the symptoms and the differences in brain structure that they believe they have found. I think it was found that London taxi drivers have a difference in the size of part of their brains, due to the large amount of work the bigger bit does remembering a spatial model of roads. So, there could be structural brain differences and the disease could still be psychosomatic.

Ah, yes, on the taxi drivers:
Significantly increased gray matter volume was found in the brains of taxi drivers compared with those of controls in only two brain regions, namely the right and the left hippocampi (Fig. 1 a and b). No differences were observed elsewhere in the brain.
Click to expand...
 
Hutan said:
Ah, yes, on the taxi drivers:
Click to expand...

Wasn't Frackowiak's taxi driver paper shown to be garbage data?

dave30th said:
how does this work
Click to expand...

It doesn't, as you say.

They don't seem to explain why the psychiatric controls are different from healthy? And if there is a difference what is the justification for putting all psychiatric patients in one group?
And so on.
 
dave30th said:
But I still don't get how they can argue there could be structure-related biomarkers and still argue that it's all a software issue. Why don't they address that contradiction? At some point, don't they need to acknowledge that something is physically wrong besides "software" and "brain network" issues?
Click to expand...
Because nobody with authority over them is calling them out on it.

This problem goes way deeper than a handful of rogue players in one small area of medicine.
 
You must log in or register to reply here.
Back
Top Bottom