LIN: Forschungsabteilungen > Akkustik, Lernen, Sprache > Unterpunkt Ebene 3 > Unterpunkt Ebene 4
Titel: LIN Layout
Associative learning has been traditionally conceived as a continuous process by which associative strength is accumulated gradually over time. However, animal studies demonstrate that in the course of associative learning, abrupt changes in learning performance, as well as rapid, learning-related transitions in brain activity can be observed from trial to trial. During detection learning in the shuttle-box, a first, important transition in learning state occurs, when animals recognize that the conditioned stimuli (CS, frequency modulated tones) are a cue associated with the unconditioned footshock (US). Although, animals do not show conditioned responses in this initial learning stage, yet, marked changes in different types of prefrontal cortical activation can be already observed (figure below). Thus, the amplitude of evoked potentials in response to the conditioned stimulus onset rapidly increases within about 5 trials of training at an individual point in the learning history of each animal. This newly emerging evoked response indicates an association between CS and US. Furthermore, it can serve as a timing signal for the evaluation of the CS-US interval, which is an important prerequisite for learning the conditioned avoidance response later in training. Concurrently, initially high theta synchronization in the intertrial-interval after the compartment change in response to the US decreases over the first trials. Theta synchronization thereby presumably reflects the working memory load related to the evaluation of the trial outcome required for the formation of the described CS-US association.
A. Example of an evoked potential in response to the conditioned stimuli recorded over the right prefrontal cortex averaged across all 60 trials of the first training session in a single animal. The first positive peak was labeled as P1 and the first negative peak as N1. B. Development of the evoked potential P1-N1 peak-to-peak amplitude (red curve), and theta-power during the intertrial-interval (blue curve) over the 20 initial trials of the first training session. Theta power (4-8 Hz) was derived from spectral analysis (1024 point Fast Fourier Transform in half overlapping hamming windows) within a 15 s time window after compartment change. All values were smoothed across 5 consecutive trials in steps of a single trial. For better comparison, smoothed values were z-standardized across trials.