If you don’t need different methods for f you could define it with keyword arguments and call it by splatting those, i.e., f(; params...) as also suggested by @brendaisy.
In general, to replicate the functionality of R requires non-standard evaluation and could be done with a macro in Julia. The package StatsPlots has a similar macro for tables which allows you to write: @df mytable :a + :b. You could check their implementation and build something similar for named tuples. Yet, given the short syntax for unpacking names tuples as shown by @albheim a macro might not be necessary here.