Episodic memory in the human prefrontal cortex

Abstract

This thesis consisting of five papers examines the relationship between haemodynamic activation in the prefrontal cortex with episodic memory function in the healthy human brain. Brain function is measured with positron emission tomography (PET) and functional magnetic resonance imaging (FMRI) and episodic memory for words as well as visuo-spatial material is examined.

The first paper consists of two experiments using either verbal or visuo-spatial material and examines the effects of practice on episodic memory in repeated cycles of encoding and retrieval. Subjects are scanned with PET during retrieval. It is found that before practice, the tasks activate fronto-parietal networks and that this activity diminishes with practice in favour of paracentral brain activity, perhaps reflecting decreasing demands on working memory.

The second paper examines incidental encoding of figurative and non-figurative drawings while processing of the drawings takes place in different levels of processing (LOP). Subjects are biased toward either ‘deep’ or ‘shallow’ by rating agreeability or graphical quality of the drawings respectively. A behavioural experiment is undertaken in the first part of the paper, establishing that the LOP effect extends to visuo-spatial material. PET scanning then shows increased activity for deep encoding compared to shallow encoding in left prefrontal, parietal and anterior temporal regions, while comparing shallow encoding to deep encoding yielded right prefrontal, parietal and posterior temporal activations.

The third paper continues to examine effects of practice in another setting. Subjects are taught a specific aid to memory, a mnemonic known as the method of loci. They are scanned with PET while encoding a list of words both with and without use of the mnemonic as well as during acquisition of the mnemonic. Use of the mnemonic is associated with increased activity in left frontal and occipito-parietal networks. Acquisition of the mnemonic compared to baseline is associated with increased activity in lateral and medial parietal cortex while activity in the left medial temporal lobe increased with practice.

In the fourth paper, FMRI is used to expand the findings from the second paper by applying the greater tem- poral resolution of event-related FMRI to separate activity for the figurative from the non-figurative drawings. Only some of the results from the second paper are replicated. In the last paper, the ‘remember’/‘know’ paradigm is used. This means that subjects are given a recognition test of previously seen words and asked to indicate the quality of a recognition experience as either ‘remembered’, meaning that the subject can recall the encoding event, or ‘known’, meaning that the subject finds the word familiar but has no recollection of the encoding event. The recognition test as well as the attribution of the recognition experience as ‘remembering’ or ‘knowing’ is scanned using event-related FMRI. Recognised words that are later ‘remembered’ compared to words that are later ‘known’ are associated with increased activity in the posterior part of the right hippocampus and comparing later ‘remembered’ words with correct rejections revealed activations in the left inferior parietal lobule and in the right middle frontal gyrus.

We conclude that practice in an episodic memory task represents a change from controlled to automatic processing; that the strong form of the hemispheric encoding retrieval asymmetry hypothesis (HERA) is not sufficient to account for the patterns of activity in the frontal cortex associated with episodic memory; that levels of processing can be shown with visuo-spatial material and that levels of processing are reflected in differently lateralised frontal and parietal activity.