How similar are motor imagery and movement?

It has been suggested that motor imagery (MI) has the basic components of real motion. This possibility was tested here in 17 healthy volunteers studied while performing or imaging a fast sequence of finger movements of progressive complexity, a fast and precise extension of the arm to touch a small circle with the tip of a pencil, a periodic repetitive flexion-extension of the index finger at a specified rate, and a velocity-regulated continuous rotary movement of the right hand. Motor sequences of 4 to 5 fingers showed a real-virtual congruency similar to that previously reported with other equivalent tests, but it decreased in the simplest sequences performed with 1 to 2 fingers. A more marked decrease of real-virtual congruency was found in the experimental paradigm aimed at producing movements with a pre-specified velocity, which was low for rhythmic movements of the index finger and practically absent in the continuous rotary movements of the hand. Present data show that the ability of MI to produce "realistic" simulations of motion is not the same for all motor tasks.