Back

Data-driven brain atrophy staging in spinocerebellar ataxia type 3

Baumesiter, H.; Schaprian, T.; Wegner, P.; Ferreira, M.; Kuegler, D.; Franca, M. C.; Rezende, T. J. R. d.; Martinez, A. R. M.; Jiang, H.; Grobe-Einsler, M.; Koyak, B.; Oender, D.; van de Warrenburg, B.; van Gaalen, J.; Durr, A.; Coarelli, G.; Synofzik, M.; Schoels, L.; Giunti, P.; Garcia-Moreno, H.; Oz, G.; Joers, J. M.; Timmann, D.; Jacobi, H.; de Vries, J.; Barker, P.; Ratai, E. M.; Reetz, K.; Infante, J.; Huebener-Schmid, J.; Klockgether, T.; ESMI MR study group, ; DELCODE/DANCER, ; Berron, D.; Faber, J.

2024-05-29 neurology
10.1101/2024.05.29.24307992 medRxiv
Show abstract

Spinocerebellar ataxia type 3 (SCA3) is the most common autosomal dominant ataxia worldwide. First targeted gene therapy trials have started, offering the intriguing scenario of preventive treatment. SCA3 is associated with progressive regional brain atrophy that starts before clinical manifestation. We aimed to identify the spatiotemporal progression pattern of brain atrophy of SCA3 with a focus on early disease stages. T1-weighted MRI scans of 300 SCA3 mutation carriers and 317 controls were analyzed. Subtype and Stage Inference (SuStaIn) was used to identify the sequence of volume loss across selected brain regions. We observed one distinct sequence of brain atrophy events in SCA3 without evidence for the existence of alternative cascades. Atrophy started in the most caudal parts of the brainstem. Almost all preataxic SCA3 mutation carriers clustered in the first atrophy stages. Certainty of sequence estimation was highest for early atrophy stages with prominent involvement of the pons and cerebellar white matter. Brain atrophy in SCA3 follows a clear and distinct sequence ascending from the lower brainstem with an early involvement of white matter. Knowledge of this sequence might support the stratification of SCA3 mutation carriers with an imminent clinical onset for early interventions.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.