Thomas T. Imhoff

Using electrical noise to enhance the ability of humans to detect subthreshold mechanical cutaneous stimuli

Stochastic resonance (SR) is a phenomenon wherein the response of a nonlinear system to a weak input signal is optimized by the presence of a particular, nonzero level of noise. Our objective was to demonstrate cross-modality SR in human sensory perception. Specifically, we were interested in testing the hypothesis that the ability of an individual to detect a subthreshold mechanical cutaneous stimulus can be significantly enhanced by introducing a particular level of electrical noise. Psychophysical experiments were performed on 11 healthy subjects. The protocol consisted of the presentation of: (a) a subthreshold mechanical stimulus plus electrical noise, or (b) no mechanical stimulus plus electrical noise. The intensity of the electrical noise was varied between trials. Each subjects ability to identify correctly the presence of the mechanical stimulus was determined as a function of the noise intensity. In 9 of the 11 subjects, the introduction of a particular level of electrical noise significantly enhanced the subjects ability to detect the subthreshold mechanical cutaneous stimulus. In 2 of the 11 subjects, the introduction of electrical noise did not significantly change the subjects ability to detect the mechanical stimulus. These findings indicate that input electrical noise can serve as a negative masker for subthreshold mechanical tactile stimuli, i.e., electrical noise can increase the detectability of weak mechanical signals. Thus, for SR-type effects to be observed in human sensory perception, the noise and stimulus need not be of the same modality. From a bioengineering and clinical standpoint, this work suggests that an electrical noise-based technique could be used to improve tactile sensation in humans when the mechanical stimulus is around or below threshold. (c) 1998 American Institute of Physics.

Enhancing aperiodic stochastic resonance through noise modulation

We show that the conventional stochastic resonance (SR) effect for aperiodic signals in a model neuron can be enhanced by modulating the intensity of the input noise (which could be introduced artificially in bioengineering applications) with either the input signal or the units output rate signal. We analyze SR enhancement theoretically and numerically. We discuss how this work provides the theoretical foundation for the development of an optimal noise-based technique for enhancing sensory function. (c) 1998 American Institute of Physics.