Figure 2
From: Normalizing brain activity across individuals using functional reference mapping

Schematic of the functional reference mapping procedure. From top to bottom: in the optical imaging recording of a fly antenna (left), c = 300 features are selected (centre), and for each of these features, the 10 most informative time points (t) are taken (responses to the first and second odorant stimulus, right). Next, signals are processed (ΔF/F, autoscaling), separately for the test odorants (right) and the r = 3 reference odorants (left). Each odorant response (test odour, or reference odour) gives a vector sized q = t * c. These are the v i local for test odorants, while the r reference odorants are assembled to the matrix P i (for this animal i). P i is used to calculate an animal-specific S i , and S i is used to remap each v i local into v i global.