Skip to main content

Table 2 Confusion matrix for the three-way classification performance using the features derived from the left-frontal region

From: Integrated space–frequency–time domain feature extraction for MEG-based Alzheimer’s disease classification

Test observation: n = 54, each class with 18 Predicted AD Predicted MCI Predicted HC
Actual AD 18 0 0
Actual MCI 0 16.5 1.5
Actual HC 0 6.1 11.9
  1. The results are obtained using the optimal feature engineering approach, i.e. the horizontal detail coefficients with Type-SD, from the gradiometer modality