Personalized MRI-based characterization of subcortical anomalies in Ataxia-Telangiectasia using deep-learning

Cerebellar atrophy is a known feature of ataxia-telangiectasia (A-T). However, basal ganglia dysfunction contributing to extrapyramidal movement disorders in A-T remains understudied. To characterize basal ganglia abnormalities in A-T using a normative self-supervised deep autoencoder trained on MRI...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	PloS one Ročník 20; číslo 8; s. e0328828
Hlavní autoři:	Saini, Catalina, Salazar-Vilches, Cristian, Blanchard, Caroline C.V., Whitehouse, William P., Parra, Denis, Dineen, Rob A., Pszczolkowski, Stefan
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	United States Public Library of Science 29.08.2025 Public Library of Science (PLoS)
Témata:	Abnormalities Adolescent Anomalies Artificial intelligence Ataxia Ataxia telangiectasia Ataxia Telangiectasia - diagnostic imaging Ataxia Telangiectasia - pathology Atrophy Basal ganglia Basal Ganglia - diagnostic imaging Biology and Life Sciences Blood flow Cerebellum Cerebellum - diagnostic imaging Cerebellum - pathology Cerebral blood flow Cerebrovascular Circulation Child Datasets Deep Learning Demographic aspects Diagnosis Diffusion coefficient Dystonia Errors Female Flow mapping Ganglia Globus pallidus Health aspects Hippocampus Humans Kinases Machine learning Magnetic resonance imaging Magnetic Resonance Imaging - methods Male Medical imaging Medicine and Health Sciences Movement disorders Neuroimaging Perfusion Putamen Reconstruction Research and Analysis Methods Research ethics Risk factors Substantia alba Substantia grisea Thalamus Chile Canada Calgary Alberta Canada
ISSN:	1932-6203, 1932-6203
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Popis
Shrnutí:	Cerebellar atrophy is a known feature of ataxia-telangiectasia (A-T). However, basal ganglia dysfunction contributing to extrapyramidal movement disorders in A-T remains understudied. To characterize basal ganglia abnormalities in A-T using a normative self-supervised deep autoencoder trained on MRI-based diffusion and perfusion features from healthy children. Mean values of apparent diffusion coefficient and cerebral blood flow perfusion maps were extracted from seven regions-of-interest: caudate, hippocampus, pallidum, putamen, thalamus, cerebellar gray matter and cerebellar white matter. A normative deep autoencoder that reconstructs these features was trained on healthy subjects. Reconstruction errors for healthy and A-T participants were computed. We used Shapley Additive Explanations (SHAP) to identify the most influential features contributing to the features' reconstruction predictions. Correlations between reconstruction errors and clinical scores in A-T patients were evaluated. Features were correctly reconstructed in controls but not A-T participants, who showed significantly higher reconstruction errors. Hippocampus, caudate and putamen diffusion, and caudate and putamen perfusion were overestimated, and cerebellar diffusion and pallidum perfusion underestimated, in participants with A-T. SHAP scores revealed that caudate, putamen, and hippocampus perfusion had the greatest influence on the reconstruction of perfusion features. In contrast, cerebellar diffusion and caudate perfusion had the greatest influence on the reconstruction of diffusion features. Exploratory analysis showed that extrapyramidal movement sub-scores from A-T participants correlated with perfusion and diffusion reconstruction errors from cerebellar and subcortical structures. Our findings suggest that pallidum, caudate, and cerebellar gray matter are potential targets for novel treatment approaches for A-T. The approach enables identification of subtle tissue anomalies at an individual level, allowing tailored approaches.
Bibliografie:	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 Competing Interests: The authors have declared that no competing interests exist. CS and CSV contributed equally to this work.
ISSN:	1932-6203 1932-6203
DOI:	10.1371/journal.pone.0328828