Investigations of the neural basis of source memory strength

Woroch, Brion S

Permalink

https://hdl.handle.net/2142/78450 Copy

Description

Title

Investigations of the neural basis of source memory strength

Author(s)

Woroch, Brion S

Issue Date

2015-04-23

Director of Research (if dissertation) or Advisor (if thesis)

Gonsalves, Brian D.

Doctoral Committee Chair(s)

Cohen, Neal J.

Committee Member(s)

• Federmeier, Kara
• Benjamin, Aaron
• Fabiani, Monica

Department of Study

Psychology

Discipline

Psychology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• functional magnetic resonance imaging (fMRI)
• Event Related Potentials (ERP)
• recognition memory
• source memory

Abstract

When an attempt is made to recognize something, we can remember different aspects of the original experience. You may have a memory of the individual items you have encountered. We may also have memories of where those items came from, the source of those memories. Source memory includes elements regarding the perceptual or conceptual context of an item, such as the person who said or wrote it, whether the item was seen inside or outside, or your internal state when the memory was encoded. Some cognitive models of source memory suggest that it is a probabilistic threshold process (Yonelinas, 2002), while others predict that it varies along a continuum of memory strength (Mickes, Wais, & Wixted, 2009; Slotnick, 2010). These models have been largely debated as part the discussion of dual-process theories of memory, leaving open cognitive neuroscience questions about brain structures involved in variations in the strength of source memory (Wixted, 2009; but see Vilberg & Rugg, 2009; Yu, Johnson, & Rugg, 2012). The dominant viewpoint in cognitive Neuroscience was that source memory did not vary in strength, leading to a lack of research into the topic. However there is evidence that this may have lead to a misrepresentation of the brain’s role in the strength of source memory (Wais, 2008) . This dissertation presents a series of experiments that explore the links between brain structures and variations in the strength of source memory. Chapter 2 uses fMRI to examine the role of the hippocampus in source memory strength. Previous dual-process oriented research suggests that a process of recollection would be necessary for source memory, but a process of familiarity would be sufficient for memory of an individual item. Brain imaging studies support the idea that that recollection is mediated by the hippocampus and familiarity by surrounding cortex. However, some researchers dispute this distinction, claiming that the hippocampus responds to the overall strength of memory, not source memory specifically (Kirwan, Wixted, & Squire, 2008; Shrager, Kirwan, & Squire, 2008; Wais, 2008). The experiment in Chapter 2 measured memory type (item and source) and the strength of both to examine whether the hippocampus responds differentially to each type of memory strength. The finding was that hippocampal activity increased with the strength of source memory, but not with item memory strength. This suggests that while the hippocampus does preferentially respond to source retrieval, it does so in a graded fashion with strength. The parietal lobes also showed a differential pattern of activity to item and source strength with different subregions involved in processing item and source memory. Dorsal regions are often associated with processing item memory, and ventral regions with source memory. This may reflect differences in bottom-up versus top-down driven attentional mechanisms (Cabeza, Ciaramelli, Olson, & Moscovitch, 2008) or difference in the accumulation of different types of mnemonic information (Vilberg & Rugg, 2008). The left frontal lobe was increasingly active for item memory strength, suggesting it plays a role of increased effortful retrieval (Henson, Shallice, & Dolan, 1999) or monitoring of source information (Mitchell & Johnson, 2009) to high confidence items. This first study left open questions about the circumstances under which other brain regions contribute to source memory strength. Chapter 3 used fMRI to examine the domain specificity of source memory strength, such as which brain regions are more active for source memory in general and which regions differed in their contribution to source strength depending on stimulus content. The retrieval of information leads to the reactivation of the brain regions that were engaged during encoding of the memory (Danker & Anderson, 2010; Rissman & Wagner, 2012; Walker, Low, Cohen, Fabiani, & Gratton, 2014), this reactivation could also contribute to source memory strength. My paradigm used pairings of words with faces or scenes during encoding, and tested memory for the associated picture with words alone. I found that the hippocampus increased its activity in a domain-general way, increasing with source memory strength to words regardless of associated stimulus type (face or place). The parahippocampal place area shows preferential activity when viewing scenes (Epstein & Kanwisher, 1998). This study found the left parahippocampal cortex to be increasingly active with confidence that a word had been previously paired with a place. Bilateral amygdala is active during the processing of face stimuli (Ishai, 2008) This study found the amygdala was increasingly active with confidence that a word was previously paired with a face. These two regions show domain-specificity in their contribution to source memory strength since they only activate to words paired with certain classes of stimuli, faces or places. Thus, while perceptual processing regions of the brain specific to faces and places showed domain-specificity, the hippocampus was involved in successful source retrieval in a domain general manner, suggesting it supports memory for all types of sources. An open question was whether the activations observed in the previous experiment represented the remembering of general categorical information diagnostic of the source question or the remembering of the specific associated stimuli. Chapter 4 examined the difference in the neural response when remembering general categorical source information or memory of specific associative information using Event Related Potentials (ERPs). During encoding, participants were shown words paired with scenes that were either natural or man-made. When shown the words again during the memory test, the participants were asked a source question, whether the associated scene was natural or manmade. Immediately following the categorical source question, they were asked to identify the specific paired scene from an array consisting of the original stimulus and two within-category lures. Thus I was able to examine ERPs reflecting source category memory retrieval separately from remembering specific associated items. Previous ERP studies have found correlations between recollection of source memory and the “LPC” component (for review see Rugg & Curran, 2007). The amplitude of the LPC has also been shown to track confidence in source memory (Woroch & Gonsalves, 2010). ERP amplitudes in the LPC time window were related to confidence in the category source judgment, but not to success at recognizing the specific associated image. Immediately following the scene category decision, subjects were given a forced choice test probing their memory for the specific associated scene. Success in that decision was related to increased amplitude late in the time window. I conclude that remembering specific information involves increases source-monitoring demands indexed by the frontal ERPs. Across all three studies, multiple brain regions and processes were involved in source memory strength, suggesting source memory is a continuous process, not simply an all or none threshold process. The fMRI data suggest the hippocampus and ventral parietal cortex track source memory in a domain-general fashion, suggesting a global binding mechanism (hippocampus), and attention to memory or memory representation (ventral parietal) processes. Domain-specific perceptual processing regions showed reactivation during retrieval that also tracked source strength, suggesting source memory strength is also affected by the richness of the reinstatement of the memory representation. The ERP results suggest that while these domain general processes, such as source retrieval indexed by the LPC, do support general source memory strength, additional executive processes, such as post-retrieval monitoring, may be necessary to support memory for specific associations at retrieval.

Graduation Semester

2015-5

Type of Resource

text

Permalink

http://hdl.handle.net/2142/78450

Copyright and License Information

Copyright 2015 Brion Woroch

Owning Collections