THE RELATIVE CONTRIBUTION OF RETINAL AND EXTRARETINAL SIGNALS IN DETERMINING THE ACCURACY OF REACHING MOVEMENTS IN NORMAL SUBJECTS AND A DEAFFERENTED PATIENT

This experiment investigated the relative extent to which different si gnals from the visuo-oculomotor system are used to improve accuracy of arm movements. Different visuo-oculomotor conditions were used to pro duce various retinal and extraretinal signals leading to a similar tar get amplitude: (a) fixating a central target while pointing to a perip heral visual target, (b) tracking a target through smooth pursuit move ment and then pointing to the target when its excursion ceased, and (c ) pointing to a target reached previously by a saccadic eye movement. The experiment was performed with a deafferented subject and control s ubjects. For the deafferented patient, the absence of proprioception p revented any comparison between internal representations of target and limb (through proprioception) positions during the arm movement. The deafferented patient's endpoint therefore provided a good estimate of the accuracy of the target coordinates used by the arm motor system. T he deafferented subject showed relatively good accuracy by producing a saccade prior to the pointing, but large overshooting in the fixation condition and undershooting in the pursuit condition. The results sug gest that the deafferented subject does use oculomotor signals to prog ram arm movement and that signals associated with fast movements of th e eyes are better for pointing accuracy than slow ramp movements. The inaccuracy of the deafferented subject when no eye movement is allowed (the condition in which the controls were the most accurate) suggests that, in this condition, a proprioceptive map is involved in which bo th the target and the arm are represented.