COMPARISON OF TRANSIENT AND STEADY‐STATE METHODS *

Transient stimuli follow each other a t sufficiently long intervals that the visual system returns to its initial state before the next stimulus occurs. Steady-state stimuli are delivered at a greater rate, so that the response to one stimulus has not died away before the next stimulus is delivered. Transient stimulation gradually changes into steady-state stimulation over a range of stimulus repetition rates. The minimum stimulation rate for steady-state stimulation is the lowest rate for which appreciable overlap occurs between the response to one stimulus and the Occurrence of the next stimulus. This rate is different for different types of stimuli. For a linear system the transient response has a fixed relationship to the steady-state response. Consequently, transient and steady-state descriptions of a linear system’s behavior are equivalent, and can be regarded as alternative formulations of the same data. However, the visual and auditory pathways commonly show several types of nonlinear behavior. For example, when successive responses run into each other they may not summate even approximately linearly. In principle, therefore, transient and steady-state stimulation can produce responses that provide complementary information about the sensory system under test. However, this possibility has so far been little exploited. At the present time, the chief practical distinction between transient and steady-state evoked potentials lies in the trade-off of speed versus information. As discussed below, recording steady-state evoked potentials can be much speedier than recording transient evoked potentials, and can be accomplished for much lower signal levels even in adverse environments. For example, mains interference is usually no problem when recording steady-state evoked potentials, but care must be taken to minimize it when recording transient evoked potentials. The price that must be paid for these advantages is that, over a given recording time, steady-state evoked potentials provide less information than transient evoked potentials. As described below, steady-state evoked potentials can be quantified in a less arbitrary manner than transient evoked potentials. Steady-state evoked potentials can be split up into a small number of harmonic frequency components, each recorded by a separate Fourier analyzer. A Fourier analyzer’s output is equivalent to two numbers (namely, amplitude and phase), and these are quite unequivocal. In contrast, describing and measuring a transient evoked potential may not be straightforward, especially when the waveform is unusual or complex (FIGURE 2 illustrates one of the many problems). In such situations different experimenters may measure transient evoked potentials quite differently, may disagree as to whether an evoked potential is normal or abnormal, and may even disagree as to whether an evoked potential is