Interestingly, both curcumin and ellagic acid are reported to have an effect as MAO inhibitors. This puts them into the same ground as selegiline and rasagiline (Azilect). Working on rats, Khatri and Juvekar write [1] that 50% of MAO-B activity is inhibited by:
Curcumin 500.46nM
Ellagic acide 400.24
Selegine 65.5

Let's see where a very rough and ready analysis of the data takes us. The levodopa equivalent dose conversion factor for sublingual selegiline is 80 and for rasagiline it is 100 [2]. Given this, and assuming that the molecular weights and bio-availability are similar, this gives a curcumin conversion factor of 80*65.5/500.46, that is about 10. Given, also, that 1mg is a therapeutic dose of rasagiline, this suggests that 10mg is, for this pathway, a therapeutic dose of curcumin. Note, unfortunately, it is not additive beyond 1mg of rasagiline, because once all the MAO-B has been inhibited, there's nothing more to be gained by increasing the dose.

Reference:

[1] "Kinetics of inhibition of monoamine oxidase using curcumin and ellagic acid"
Dharmendra Kumar Khatri, Archana Ramesh Juvekar
Pharmacognosy Magazine, 11 May, 2016
Kinetics of inhibition of monoamine oxidase using curcumin and ellagic acid Khatri DK, Juvekar AR - Phcog Mag

[2] Parkinson's Disease Measurement: PwP, surveys, trials, analysis

John