Autismus: Gehirne je nach Geschlecht anders
Die Organisation des Gehirns ist bei Jungen und Mädchen mit Autismus unterschiedlich, wie Forscher der Stanford University School of Medicine http://med.stanford.edu berichten. Die Unterschiede, die durch Analyse von hunderten Gehirn-Scans mittels Techniken der Künstlichen Intelligenz (KI) festgestellt wurden, waren für Autismus spezifisch und konnten bei sich normal entwickelnden Kindern nicht festgestellt werden. Autismus wird bei vier Mal so vielen Jungen diagnostiziert als bei Mädchen. Der Großteil der Autismusforschung hat sich daher auf männliche Personen konzentriert.
...
Von den Kindern mit Autismus verfügten die Mädchen in verschiedenen Gehirnzentren wie bei der Motorik, Sprache und der visuellräumlichen Aufmerksamkeit, über unterschiedliche Muster bei der Konnektivität. Die Unterschiede in einer Gruppe von motorischen Bereichen wie dem primären Motorcortex, dem ergänzenden motorischen Bereich und seitlichem okzipitalen Kortex sowie dem Gyrus temporalis medius und superior waren am größten. Bei Mädchen mit Autismus standen die Unterschiede in den motorischen Zentren mit der Schwere ihrer motorischen Symptome in Zusammenhang. Mädchen, deren Gehirnmuster denen von betroffenen Jungen am ähnlichsten waren, neigten daher dazu, unter den am stärksten ausgeprägten motorischen Symptomen zu leiden.

Link zum Artikel bei Pressetext

Deep learning identifies robust gender differences in functional brain organization and their dissociable links to clinical symptoms in autism

Discussion
Main findings
By using one of the largest functional brain imaging cohorts of females and males with ASD and leveraging exciting recent advances in XAI, we examined neurobiological gender differences in ASD. To our knowledge, this is the first use of a XAI-based approach for uncovering robust neurobiological gender differences in ASD. Our XAI-based approach is also a significant advance over previous approaches to find gender differences in ASD using functional brain imaging data (see Supplementary Discussion for details). Our novel XAI-based stDNN model, which uses spatiotemporal convolution on fMRI data to distinguish between groups, achieved consistently high classification accuracy in distinguishing between females and males with ASD. Notably, the stDNN model trained to distinguish between females and males with ASD could not distinguish between neurotypical females and males, suggesting that there are gender differences in the functional brain organisation in ASD and these gender differences are different from normative gender differences.

Brain features associated with motor, language and visuospatial attentional systems reliably distinguished females with ASD from males with ASD. Crucially, these results were observed in the large multisite ABIDE/Stanford cohort and replicated in a fully independent CMI-HBN cohort. Furthermore, brain features associated with the primary motor cortex node of the motor network emerged as a robust predictor of the severity of RRBs in females with ASD but not in males with ASD.
..
The second main finding of our study is the identification of functional brain features that distinguish between females and males with ASD. Our stDNN-based integrated gradients analysis identified the primary motor cortex and supplementary motor area, which anchor the motor network, as brain areas whose dynamical functional properties most clearly distinguished between females and males with ASD in both cohorts. Aberrancies in the extended motor network have been consistently reported in ASD, when compared with neurotypical individuals.
...
In both cohorts, our analysis also identified bilateral middle and superior temporal gyri, which anchor the language system, as brain areas whose features clearly distinguished between females and males with ASD. Impairments in temporal cortex areas associated with language processing are a prominent feature of ASD. Language systems anchored in the middle and superior temporal gyri facilitate speech processing and semantic comprehension, processes known to be impaired in ASD.

Surprisingly, our analysis also identified the dorsal parietal cortex as a brain area whose features distinguished between females and males with ASD in both cohorts. Specific loci included the superior parietal lobule and the adjoining posterior human intraparietal area 3 (hIP3) subdivision of the intraparietal sulcus, which play a crucial role in visuospatial attention.
Moreover, these parietal areas, encompassing the dorsal visual stream pathway, have strong connections with the lateral occipital gyrus whose features also distinguished between females and males with ASD. Together, these regions constitute key elements of the dorsal and ventral visual pathways involved in attending to the location of objects in space. This finding of gender-related differences is noteworthy because of conflicting evidence in the extant literature about co-occurring parietal and lateral occipital cortex impairments in ASD.
Taken together, these results identify gender differences in motor and language systems that are known to be impaired in the disorder as well as dorsal parietal and lateral occipital cortex regions that have not been reported previously in the literature as being affected in ASD. Interestingly, although default mode network and salience network aberrancies have been consistently reported in ASD,gender differences were not reliably observed in these regions, suggesting that aberrancies in the default mode network and salience network may be common to females and males with ASD.
...
A potential avenue for investigation might be to examine the relationship between RRBs in males with ASD and the functional organisation of other components of the motor network, including the basal ganglia and cerebellum, which we recently found to be predictive of repetitive motor behaviours in a predominantly male ASD sample.

Link zur Studie bei Cambridge