Hi MNE users,
When comparing the goodness of fit for dipoles across different software (MNE vs Brainstorm), I found that possibly the way they both define the term Goodness of fit is different. My fitted dipole on MNE was actually closer to the actual current source, however, it shows a Goodness of fit of only 39%, the dipole fitting approach on brainstorm yields a Goodness of fit of 94% but it is away from the real source compared to the MNE’s fitted dipole.
I found this across the Brainstorm website. But my question is what exactly is the formula for residual variance and what does it depend upon?
What approach does MNE take? How does it define Goodness of fit? Is it 1- residual variance or just the residual variance?
