Classification of visual sensations generated electrically in the visual field of the blind

Within the framework of the OPTIVIP project, an optic nerve based visual prosthesis is being developed in order to restore partial vision to the blind. In this paper, we concentrate on the classification problem of visual sensations generated by multiple electrical stimulations. We propose to use probabilistic neural networks in order to perform Bayesian classification. Statistical sampling techniques are utilized in order to reduce the bias on the estimated performances and assess the sensitivity of the method. In noisy environment the Parzen window estimator seems more reliable than finite Gaussian mixtures.