Abstract
Abstract: :
Purpose: This study was undertaken in order to evaluate the benefits to the user of a visual prosthesis based on the electrical stimulation of the optic nerve (Veraart et al., ARVO 1999). The implanted volunteer, a completely blind retinitis pigmentosa patient, executed several tasks simulating daily life object localisation, discrimination, and grasping. Methods: The prosthesis was composed of a head worn video camera connected to an external processor which generated the optic nerve stimulation parameters (Delbeke et al., this meeting). The nerve was stimulated whenever a part of the processed image of an examined object matched an elicitable phosphene within the remaining visual field of the volunteer. The volunteer sat in front of a table divided in 9 rectangular zones. Six objects were used : a large bottle, a small bottle, a cup, a CD box, a knife, and a toothpaste box. The tasks consisted in the localisation, the discrimination, and the grasping of a given object either alone, or among an arrangement of 3. Scores and task duration were recorded. A portable equipment has also been prepared to investigate performances in mobility. Results: After several months of training with interleaved evaluations the volunteer's score reached 100 % of success : (1) at the 3rd evaluation to localise a single object, in about 15 s, and to grasp it in about 5 s, (2) at the 5th evaluation to discriminate a single object, in about 40 s, and (3) at the 1st evaluation to grasp a specific object among 3, in about 60 s. Conclusion: These results show that in our experimental conditions the implanted volunteer proved able to correctly localise, discriminate, and grasp an object among others. The task duration remained substantial with the present algorithm. However a higher rate of visual information transmission is eventually expected with a more efficient algorithm.
Keywords: electrophysiology: non-clinical • learning • space and scene perception