My intuition says that you want to apply a known pressure and measure deflection. At least that way you should have a fairly failsafe probe. I would worry about a stepper driven mechanism pressing into human tissue (i.e. me) in case it doesn't stop, even with an emergency cut off switch.

Instead of a linear voice coil, you could use a DC motor to drive a rack and pinion probe mechanism. Assuming you can get your friction down, the relationship between motor current and force is pretty relaible. You could measure deflection using an encoder on the motor shaft.