The medial prefrontal cortex (mPFC) is part of a core network that not only supports the recollection of past episodes but also the imagining of prospective events (e.g., meeting a person at a particular place). Here, we use such episodic simulation and fMRI to provide insights into the exact functions supported by the mPFC. In the first part, RMWJB reports graph analytical evidence that the ventral mPFC acts as a hub that coordinates whole-brain connectivity. This region thus seems to support episodic simulation by effectively integrating distributed information (e.g., about the people and places featuring in an imagined event). In the second part, PCP uses representational-similarity analysis to examine the nature of the representations in the mPFC. Specifically, he tests the hypothesis that this region codes for affective associative representations of our environment. Together, the two parts will shed light on the functions supported by the mPFC by highlighting both, its internal representations and its influence on the brain’s overall functional connectivity. [mehr]

[mehr]

[mehr]

The human medial temporal lobe (MTL) is essential for episodic memory, but the roles of its individual subregions remain a matter of debate. Anatomical and functional evidence suggests a division of labour based on stimulus domain: The MTL cortex may process domain-specific aspects of memory, with the anterior MTL cortex [perirhinal and anterolateral entorhinal cortices (PRC, alEC)] supporting object-related memory, and the posterior MTL cortex [parahippocampal and posteriormedial entorhinal cortices (PHC, pmEC)] supporting spatial memory. The hippocampus, on the other hand, may contribute to memory in a domain-general fashion. Here, I present data from three studies. Studies 1 and 2 investigated the domain-specificity of episodic memory using 3T fMRI. During incidental encoding (study 1) and associative recall (study 2), PRC and PHC contributed to object and scene memory, respectively. Study 3 utilised the ultra-high spatial resolution afforded by 7T fMRI to investigate the role of alEC, pmEC, and hippocampal subfields during associative encoding and retrieval of objects and scenes. Here, I will present preliminary results and highlight methodological challenges. In sum, these studies emphasise the role of the human MTL in episodic memory, with subregional differences that depend on stimulus domain. [mehr]

Learning often takes place in environments, in which it is impossible to exactly know current and future outcomes. To successfully behave in such uncertain environments, humans have to learn appropriate beliefs from past experiences that can be used to predict desirable and undesirable outcomes. Drawing on optimal inference models and behavioural learning tasks, I will illustrate how learning under uncertainty should be regulated from a normative perspective and how learning deficits may emerge from deviations from these computations. I will show how human participants learn in the face of perceptual uncertainty and to which extent the ability to adjust learning in dynamically changing environments differs between age groups across the lifespan. Moreover, I will explore the possibility that the intricate computations to optimally adjust learning may often be simplified by resorting to heuristic strategies that are guided by previous choices. Finally, I will discuss some future directions that follow from these results. [mehr]

[mehr]

[mehr]

[mehr]

[mehr]

In recent years, the Max Planck Research Group: Adaptive Memory has examined how we use our general knowledge about the world for the recollection of the past and the simulation of the future. In this colloquium, we will showcase some of our recent work-in-progress. Heidrun Schultz will first examine how we use our pre-existing knowledge to augment the retrieval of specific episodic memories. Towards this end, she gauges how information pertaining to pre-existing knowledge is represented in the medial temporal lobes versus prefrontal cortex. Mark Lauckner will then highlight the contribution of memory schemas to episodic memory and future simulation. This work combines natural language processing of a large data set with examinations of patients with focal brain lesions. Finally, Aroma Dabas will demonstrate that we learn from simulated experiences much the same way as we learn from real experiences. Specifically, she will show that simulation-based learning is governed by the same computational algorithm and implemented by a similar neural mechanism. [mehr]

[mehr]

[mehr]