Theory of Visual Attention (TVA) applied to mice in the 5-choice serial reaction time task

Abstract

Rationale: The 5-choice serial reaction time task (5-CSRTT) is widely used to measure rodent attentional functions. In humans, many attention studies in healthy and clinical populations have used testing based on Bundesen's Theory of Visual Attention (TVA) to estimate visual processing speeds and other parameters of attentional capacity.

Objectives: We aimed to bridge these research fields by modifying the 5-CSRTT's design and by mathematically modelling data to derive attentional parameters analogous to human TVA-based measures.

Methods: C57BL/6 mice were tested in two 1-h sessions on consecutive days with a version of the 5-CSRTT where stimulus duration (SD) probe length was varied based on information from previous TVA studies. Thereafter, a scopolamine hydrobromide (HBr; 0.125 or 0.25 mg/kg) pharmacological challenge was undertaken, using a Latin square design. Mean score values were modelled using a new three-parameter version of TVA to obtain estimates of visual processing speeds, visual thresholds and motor response baselines in each mouse.

Results: The parameter estimates for each animal were reliable across sessions, showing that the data were stable enough to support analysis on an individual level. Scopolamine HBr dose-dependently reduced 5-CSRTT attentional performance while also increasing reward collection latency at the highest dose. Upon TVA modelling, scopolamine HBr significantly reduced visual processing speed at both doses, while having less pronounced effects on visual thresholds and motor response baselines.

Conclusions: This study shows for the first time how 5-CSRTT performance in mice can be mathematically modelled to yield estimates of attentional capacity that are directly comparable to estimates from human studies.

Keywords: 5-choice serial reaction time task; Attention; Attention deficit/hyperactivity disorder; Mouse behaviour; Theory of Visual Attention; Touchscreen operant chamber; Translatable animal models.