Activation of single cells in cat visual cortex by electrical stimulation of the cortical surface

Electrical stimulation of the brain and the development of cortical visual prostheses: An historical perspective

Rapid advances are occurring in neural engineering, bionics and the brain-computer interface. These milestones have been underpinned by staggering advances in micro-electronics, computing, and wireless technology in the last three decades. Several cortically-based visual prosthetic devices are currently being developed, but pioneering advances with early implants were achieved by Brindley followed by Dobelle in the 1960s and 1970s. We have reviewed these discoveries within the historical context of the medical uses of electricity including attempts to cure blindness, the discovery of the visual cortex, and opportunities for cortex stimulation experiments during neurosurgery. Further advances were made possible with improvements in electrode design, greater understanding of cortical electrophysiology and miniaturisation of electronic components. Human trials of a new generation of prototype cortical visual prostheses for the blind are imminent. This article is part of a Special Issue entitled Hold Item.

Excitation of visual cortex neurons by local intracortical microstimulation

The threshold current required for the excitation of visual cortex neurons in the vicinity (approximately 1 mm) of an intracortical stimulating electrode was measured as a function of the stimulus pulse duration in the anesthetized cat. For cortical neurons with latencies of activation from 0.4 to 3.4 ms and for stimulus pulse durations from 0.02 to 0.7 ms, the threshold current for all neurons tested decreased in an exponential fashion as the pulse width was increased. Rheobase current values (ampere-threshold) were 1.2 to 516 muA (mean 160 +/- 24 muA, N = 24) and chronaxie values were 0.07 to 0.79 ms (mean 0.217 +/- 0.036 ms, N = 24). When the quantity of charge required for neuronal excitation was calculated, a quasilinear relationship was found between threshold charge and stimulus pulse width. The minimum threshold charge (coulomb-threshold) occurred for the briefest pulse widths tested and were 2 to 86 nC (mean 36.4 +/- 4.4 nC, N = 24). When the pulse energy index was calculated (threshold current squared multiplied by the pulse width), the minimum pulse energy capable of generating an evoked response (a single action potential) occurred when the pulse width was approximately 80% greater than the chronaxie. These studies demonstrate that the predictions derived from A. V. Hill's classical theory of nerve excitation are to a first approximation obeyed by visual cortex neurons. For the three parameters analyzed as a function of stimulus pulse width, the pulse current is minimized at long pulse durations, the pulse charge is minimized at short pulse durations, and the pulse energy is minimized at pulse widths of intermediate value.

Intracortical microstimulation of neurons in the visual cortex of the cat

The response of visual cortex neurons to local intracortical microstimulation was measured in the anesthetized cat. When the recording microelectrode was very close (about 20 micrometers) to the tip of the stimulating electrode, threshold currents as low as 10 micro A were capable of firing neurons. Over a 20-fold range in distance from the site of stimulation, an 80-fold increase in threshold current was observed. The mean latency of activation for 30 neurons tested with intracortical stimulation was 2.88 +/- 0.45 msec. The majority of these cells were probably synaptically activated. The mean threshold current for these neurons was 0.55 +/- 0.12 mA (N = 30). These values were significantly smaller than the thresholds found previously when stimulating electrodes were located on the pia-arachnoid surface of the visual cortex.