Neural code metrics: Analysis and application to the assessment of neural models

Pulmonary stretch receptor spike time precision increases with lung inflation amplitude and airway smooth muscle tension

Slowly adapting pulmonary stretch receptors (SARs) respond to different lung inflation volumes with distinct spike counts and patterns, conveying information regarding the rate and depth of breathing to the cardiovascular and respiratory control systems. Previous studies demonstrated that SARs respond to repetitions of the same lung inflation faithfully, suggesting the possibility of modeling an SAR's discrete response pattern to a stimulus using a statistically based method. Urethane-anesthetized rabbit SAR spike trains were recorded in response to repeated 9-, 12-, and 15-ml lung inflations, and used to construct model spike trains using K-means clustering. Analysis of the statistics of the responses to different lung inflation volumes revealed that SARs fire with more temporal precision in response to larger lung inflations, because the standard deviations of real spikes clustered around the modeled spike times of responses to 15-ml stimuli were smaller than those produced by 12 or 9 ml, even at the same absolute firing frequencies. This implied that the mechanical coupling of SAR endings with pulmonary tissue is critical in determining spike time reliability. To test this, we collected SAR responses during bronchial constriction, compared them with those produced by the same SARs under normal airway resistance, and found that their firing reliability improved during bronchial constriction. These results suggest that airway distension and mechanical coupling of the receptor endings with the airway wall (partially determined by smooth muscle tone) are important determinants of SAR spike time reliability.