Table 2 Summary of common performance metrics for the different classification models

From: A functional neuroimaging biomarker of mild cognitive impairment using TD-fNIRS

Metric

Accuracy

AUC

Sensitivity (recall)

Specificity

Precision

NPV

Model Input

Survey only ADCS-ADL-MCI

0.73

0.76

0.62

0.84

0.79

0.69

Behavior only

0.71

0.73

0.70

0.73

0.71

0.71

Brain only

0.64

0.72

0.64

0.65

0.64

0.65

Survey + Behavior

0.71

0.79

0.64

0.78

0.74

0.69

Survey + Behavior + Brain

0.85

0.92

0.80

0.90

0.89

0.82

Brain + Behavior

0.84

0.90

0.78

0.90

0.89

0.80

  1. Each row indicates a model with the described input data. Behavior data consists of the performance metrics from both the N-Back and VFT tasks. The brain data consists of GLM features from the two cognitive tasks.
Back to article page