HIGHLIGHTS
SUMMARY
The authors report the performance of individual features extracted from recorded spoken responses and their transcripts and classifiers built by combining the features at distinguishing healthy controls (HC) from study participants aMCI or naMCI. In three separate analyses, cross-validated elastic-net logistic regression is used to select a subset of numerical features and then form a weighted combination of the selected features that optimally separate the HC group from either aMCI, naMCI, or combined MCI (aMCI+naMCI) cases. In this study, analyses suggest that language features based on picture description transcripts may be . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.