Structural MRI and functional connectivity features predict current clinical status and persistence behavior in prescription opioid users.

This yielded numerically worse classification accuracy than the “sMRI only” classifier, and failed to reach statistical significance (see Fig. 2 ): 57.1% mean accuracy, 61.9% POUD sensitivity, 52.4% control specificity; p = .105, null mean = 49.8%.

also in 5,197 other papers

Comparison of the numerical differences in prediction accuracy across feature sets revealed a similar pattern to the ML classification results: prediction accuracy was clearly non-significant for the “sMRI subcortex” model, marginally non-significant for the “sMRI only” model and clearly significant for the “restFC only” model 5 .

also in 481 other papers