Separate mesocortical and mesolimbic pathways encode effort and reward learning signals.

Area of Science:

• Neuroscience
• Cognitive Science
• Decision Science

Background:

• Optimal decision-making requires learning choice option features, including effort and reward.
• The mesolimbic network is known for reward learning, but its role in effort learning is unclear.

Purpose of the Study:

• To investigate whether effort learning utilizes the same neural network as reward learning.
• To identify the brain regions involved in encoding effort and reward prediction errors (PEs).

Main Methods:

• Computational functional Magnetic Resonance Imaging (fMRI) was employed.
• Analysis focused on the parallel encoding of effort and reward prediction errors.

Main Results:

• Effort PEs were encoded in the dorsomedial prefrontal cortex, while reward PEs were found in the ventral striatum.
• Both effort and reward PEs shared a common mesencephalic origin in distinct dopaminergic midbrain regions.
• Reward and effort expectations were integrated in the ventral striatum during action anticipation.

Conclusions:

• Motivationally relevant stimulus features, like effort and reward, are learned via parallel dopaminergic pathways.
• An integrated utility signal for choice is formed through the integration of these parallel signals.