Back

Data-driven algorithm for the diagnosis of behavioral variant frontotemporal dementia

Manera, A.; Dadar, M.; van Swieten, J.; Borroni, B.; Sanchez-Valle, R.; Moreno, F.; LaForce, R.; Graff, C.; Synofzik, M.; Galimberti, D.; Rowe, J.; Masellis, M.; Tartaglia, M. C.; Finger, E.; Vandenberghe, R.; de Mendonca, A.; Tagliavini, F.; Santana, I.; Butler, C.; Gerhard, A.; Danek, A.; Levin, J.; Otto, M.; Frisoni, G.; Ghidoni, R.; Sorbi, S.; Rohrer, J. D.; Ducharme, S.; Collins, D. L.

2019-12-20 neuroscience
10.1101/2019.12.19.883462 bioRxiv
Show abstract

INTRODUCTIONBrain structural imaging is paramount for the diagnosis of behavioral variant of frontotemporal dementia (bvFTD), but it has low sensitivity leading to erroneous or late diagnosis. METHODSA total of 515 subjects from two different bvFTD databases (training and validation cohorts) were included to perform voxel-wise deformation-based morphometry analysis to identify regions with significant differences between bvFTD and controls. A random forest classifier was used to individually predict bvFTD from morphometric differences in isolation and together with bedside cognitive scores. RESULTSAverage ten-fold cross-validation accuracy was 89% (82% sensitivity, 93% specificity) using only MRI and 94% (89% sensitivity, 98% specificity) with the addition of semantic fluency. In a separate validation cohort of genetically confirmed bvFTD, accuracy was 88% (81% sensitivity, 92% specificity) with MRI and 91% (79% sensitivity, 96% specificity) with added cognitive scores. DISCUSSIONThe random forest classifier developed can accurately predict bvFTD at the individual subject level.

Matching journals

The top 9 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.