STIMULUS REPETITION EFFECTS ON PICTURE RECOGNITION AND EVENT RELATED POTENTIALS By J. L. Sanguinetti

A paper submitted in partial fulfillment of the requirements of the Honors Program in the Department of Psychology.

Examining Committee:

Approved By:

_________________________________

_________________________ Julian Keith, PhD Faculty Supervisor

_________________________________ _________________________________ _________________________________ Honors Council Representative

__________________________ Department Chair

_________________________________ Director of the Honors Scholars Program

University of North Carolina Wilmington Wilmington, North Carolina May 2007

2 In loving dedication to Mom, Dad, Jeanne, and Kramer

3 Contents 1

2

3

4

Introduction …………………………………………………………… 1.1 Electroencephalography……………………………………………. 1.1.1 Origins of the EEG………………………………………. 1.1.2 The Event-Related Potential Technique ………………… 1.1.3 Components …………………………………………. 1.1.4 Source Estimation …………………………………… 1.2 Recognition Memory …………………………………………... 1.2.1 Prior Occurrence …………………………………….. 1.2.2 Theoretical Framework ……………………………… 1.2.2.1 Animal Models …………………………….. 1.2.2.2 Lesion Studies ……………………………… 1.2.2.3 ERP and Dual Process Models …………….. 1.2.2.4 fMRI Evidence …………………………….. 1.3 Repetition and the Brain …………………………………….…. 1.3.2 ERP Repetition Effect ……………………………….. 1.3.3 ERP ‘Old/New’ Effect ……………………………….. 1.3.4 Repetition Suppression ………………………………. 1.3.5 The Current Study ……………………………………. Materials and Methods ………………………………………………….. 2.1 Subjects ………………………………………………………… 2.2 Equipment ……………………………………………………… 2.2.1 Electrode Layout ……………………………………… 2.3 Stimuli ………………………………………………………….. 2.4 Procedure ……………………………………………………….. 2.4.1 Experiment 1 ………………………………………….. 2.4.2 Experiment 2 ………………………………………….. Results …………………………………………………………………… 3.1 Experiment 2…………………………………………………….. 3.2 Visual Results …………………………………………………… 3.2.1 Topographic Maps …………………………………….. 3.2.2 Grand Averages ……………………………………….. 3.3 Source Estimation and Repetition Effects …………………………. 3.3.2 Amplitude Differences …………………………………. 3.3.3 Peak Latency Differences ……………………………… Discussion ……………………………………………………………….. 4.1 Findings …………………………………………………………... 4.2 Component Differences …………………………………………... 4.3 Problems ………………………………………………………….. 4.4 Future Research …………………………………………………… 4.5 Conclusion …………………………………………………………

6 6 8 11 14 21 21 21 23 26 27 29 30 31 32 33 34 36 38 38 40 42 44 46 46 52 53 53 56 56 59 71 71 77 79 79 80 81 82 82

References ………………………………………………………………..

83

4

Abstract In this study, we used state-of-the-art source estimation methods and event related potentials (ERP) to investigate the timing and amplitude of electrophysiological activity in four brain regions of interest (ROI) during a picture recognition memory task. When stimuli are repeated, subjects recognize them faster (as indexed by speeded reaction time). In the current experiment, subjects viewed pictures of animals in a study/test paradigm while continuous EEG was recorded. Pictures were repeated once, three times, or ten times in study. We believe that the same experimental factors that affect recognition will also affect the brain regions involved in recognition processes which will be revealed by a greater signal strength and decreased signal latency at regions implicated in recognition. This hypothesis was verified; source estimation analysis reveals that repetition of visual stimuli increases amplitude signal and amplitude latency from particular regions of interest within the brain.

5 Acknowledgements I would like to express my gratitude to my friends and family who have provided the social support on which I depend. Special thanks to my immediate family—Mom, Dad, Jeanne, Kramer, Grandmother, Grandfather, Bobby and Mimi—for providing your unconditional love. To Aneeka for having patience a monk would envy. To the Turrisi family (including the Cat) for providing intellectual stimulation, tea and home cooked meals. This work would never have been possible without the endless generosity of Lloyd Smith of Cortech Solutions, LLC. You not only provided the electroencephalogram and other equipment, but you also provided your time and energy and played the invisible hand in guiding this project to completion. I am indebted to you and only hope I have the opportunity to provide such generosity in the world. Many thanks to my committee for taking the time to read and critique my work, especially Dr. Cohen who allowed me to tinker in his lab when I was an EEG novice and for providing his ideas during the conception of this project. Finally, to my advisor Dr. Keith, who’s wit and humor made the hardest work I have ever done the most enjoyable. You have taken a naive college student and turned him into a naive adult, but with the skills and confidence to find his way. You stuck with me throughout our setbacks, and have been not only the best mentor imaginable, but also a true friend.

6

Chapter 1 Introduction This introduction is divided into three sections. The first section provides a theoretical overview of the electroencephalogram. This is done to explore the relationship between the extracranial electrical activity measured by scalp electrodes and the underlying bioelectrical generators in the brain. The second and third sections relate directly to the current study by summarizing what is known about recognition memory and the effects of stimulus repetition on recognition memory and the brain electrophysiology, respectively. 1.1

Electroencephalography The electroencephalogram (EEG) is a neurophysiological tool for measuring

electrical activity of the brain by means of electrodes placed on the scalp. These extracranial potentials are produced by electrical currents emanating from underlying neural populations in the brain. Unlike other imagining techniques (e.g. functional Magnetic Resonance Imaging [fMRI], Positron Emission Tomography), the EEG is continuous (i.e. in real time and ongoing). In its most simplistic form, when two electrodes are attached to the surface of the scalp and connected to an amplifier, the output will be variations in voltage over time (Rugg, 1995). Due to the spreading of electrical activity of the skull and underlying tissue, the EEG readout will always be the potential difference between at least two electrodes, with typical amplitude ranges of 100 to 100 mV, and meaningful frequencies between 0.1 and 55Hz (Rugg, 1995).

7

Figure 1. Two electrodes are detecting electrical activity on the scalp, taking the difference in the voltage at each electrode, and displaying the variation over time.

8 1.1.1

Origins of the EEG Information is transmitted throughout the brain and nervous system via the

propagation of electrochemical gradients along neuronal cell assemblies. This passage of currents within and between neurons naturally creates electrical signals that can be detected outside of the brain (by the EEG). Thus, monitoring extracellular potentials is a valid way of non-invasively measuring activity of the underlying tissue in the brain. It is widely accepted that the bioelectric potentials recorded at the scalp are generated within the brain; however, the relationship between what is going in the brain and what we observe at the scalp is not fully understood (Rugg, 1995; Nunez, 1981). These potentials are generated by the sum of the excitatory and inhibitory post-synaptic potentials (EPSPs and IPSPs) of neurons. These electrical fields are the sum of a sizable population of neurons (on the order of 10,000 neurons). Due to the nature of electromagnetism, the EEG can only measure potentials from groups of cells that are aligned perpendicular to the surface of the cortex (Zani and Proverbio, 2002). Neurons that have the necessary geometric configuration to produce fields that can be measured at the scalp do so because their individual electric fields summate to yield a net dipolar field (a field with positive and negative charges between which current flows)—known as an ‘open field’ (Rugg, 1995). It must not, however, be thought the electrical source is directly underneath the electrode. Even if the electrical activity originates at one point in the brain, the resulting EEG can be measured throughout the scalp. A specific neural generator in the brain produces a unique scalp voltage topography (the EEG), but an infinite number of different cortical source configurations

9 can produce the same scalp distribution—known as the inverse problem (discussed below) (Slotnick, 2005).

10

Figure 2. Continuous EEG from 16 electrodes. Figure 2. Continuous EEG as displayed on a typical computer screen. Show here are 16 electrodes.

11 1.1.2

The Event-Related Potential Technique Raw EEG data from individual subjects reflects “spontaneous” fluctuations of

electrical activity in the brain, and is meaningless when trying to understand underlying cognitive processes. However, researchers who study specific neural mechanisms involved in cognition (e.g. memory, attention, etc.) have devised methods to extract informative brain patterns from the raw data. Subjects are presented a series of stimuli that can be auditory or visual while the EEG is recording. In post analysis software, an epoch of the EEG that is time-locked to individual stimulus can be defined. For instance, if researchers are interested in semantic processing in the brain, they could present different words to the subject via a computer. In the software, an epoch of time can be defined around the presentation of each stimulus (e.g., 200 ms before to 1000 ms after a marker for the word in the raw EEG.) If there are voltage changes within this epoch that are specifically related to the brain’s response to the stimuli, then an event-related potential (ERP) has been found. It can be inferred that if there are similar ERPs across subjects, there is similar brain activity (and cognitive processes) across subjects. Relative to positron emission tomography (PET) and other imaging techniques, ERPs have a resolution down to the millisecond compared to seconds for the former. Equally important, ERP are reliably sensitive in detecting functional changes in brain activity. Thanks to these advantages, the event related potential technique is able to uncover steps in sensory-cognitive information processing that is occurring rapidly in the brain (Zani and Proverbio, 2002). This can help to uncover the functional organization,

12 and timing of activation, of the anatomically distributed functional systems that can be correlated with the findings of imaging studies.

13

Figure 3. A simplified model of the ERP technique for a visual word paradigm. Each time the subject views the word ‘dog’ a marker is placed in the raw EEG data. Later, the signals are averaged in the computer software and the ERP signal emerges from the background EEG.

14 1.1.3 Components In light of the discussion above, defining an ERP component might seem like an easy task. After extracting the signal using the technique described above, all one needs to do is to focus on the peak or trough of the resulting waveform and this becomes the component of interest. Measurement would be as straight forward as determining the amplitude and latency of the component in relation to some other feature of the waveform or to the baseline. There are at least two major problems to this simple approach to component definition (Donchin, et al. 1978). First, and probably the greatest impediment, is ‘component overlap’ (Picton & Struss, 1980). The waveform we observe by measuring voltage fluctuations at the scalp may be generated by several different sources in the brain. Since the brain is a conductive medium (as noted above), activity could be generated in spatial location and propagating to another where it is detected. In other words, the single voltage we measure at one location may very well be a reflection of the summation of many sources in the brain. One consequence of this is that a component (say at 200ms) may actually reflect the summation of two neural generates who were maximally active before and after that time independently, but whose open fields summated at the time recorded (Rugg, 1995). The second question is more theoretical. Some researchers such as Naatanen and Piction (1987) adopt what is called the ‘physiological approach’ in which they define the ERP component by its anatomical source within the brain. Accordingly, we must have a method of making the contributing sources unambiguous if we are going to measure particular ERP components. Other ERP researchers (see Donchin, 1979) argue that an ERP component should be defined in terms of the information processing operation with

15 which it is correlated. This is known as the ‘functional’ approach. The component is defined in terms of the cognitive function that is linked to the underlying brain system whose activity is recorded at the scalp. It is probably most realistic to adopt a mixture of both approaches (Rugg,1995). P100 The P100 is a typical sensory component found all sensory modalities. It is associated with a series of deflections in the ERP that are related to the transmission of sensory information from the peripheral sensory system to the cortex and/or the arrival of that information in the cortex (Luck, 2004). The P100 (or P1) reflects the processing of visual stimuli in the visual cortex. It is somewhat of a later sensory component than other modalities (e.g., auditory which can be on the order of 10s of ms) because the neurons in the sensory relay nuclei (e.g., the lateral geniculate nucleus) are configured in such a way they create a closed field that does not reach the scalp (Rugg, 1995). N200 About 200 ms after the presentation of auditory and visual stimuli, a negative component often becomes evident. The necessary condition to elicit this component is that the event must physically deviate in some way from the prevailing context; this is referred to as ‘mismatch negativity’ (Pritchard, 1991). In a series of now classical experiments Naatanen (1978) showed that improbable events (tones that deviated from a norm) elicited a negative component about 200 ms after presentation. P300 The P300 is the most studied ERP component. This is due in part to its side (5-20 mV) and because how easy it is to elicit the component. The standard paradigm for the P300 is

16 known as the ‘oddball paradigm.’ A series of events are presented to a subject, and the events are composed of two classes. One type of event is occurs less frequently than the other—hence the name oddball. Subjects are usually required to respond in some way to the rarer of the two events. The ERP components for the oddball stimuli usually consist of a positive deflection, that is maximal over the parietal/central area, that has a latency of 300 ms up to 900 ms (Rugg, 1995). Coles and Donchin (1988) have proposed that this effect reflects a process of context of memory updating by modifying the current model for the environment due to the incoming information.

17

Figure 4. P300 wave elicited in a typical “odd-ball paradigm.” The red line is the potential elicited from odd stimulus, the blue represents the background stimuli. (The colored circles are topographic head maps that represent the relative electrical activity on the scalp). Note: negative is up (image from Krigolson, Olav; http://web.uvic.ca/psyc/braincoglab/krigolson_res.html, with permission).

18 N400 The N400 component is sensitive to deviance to abstract attributes of the eliciting stimulus, such as meaning. Kutlas and Hillyard (1980) were the first to observe the N400 in a now classic paradigm. Subjects were required to read a series of sentences comprised of about seven words. Some sentences ended in semantically inappropriate but syntactically correct words, others with the final word printed in a larger font size than the preceding words, while other sentences were normal (they ended with appropriate words and the font was kept constant). Semantically deviant final words elicited a negative deflection with a latency of about 400 ms (the N400). Physically deviant words elicited the classical P300 (latency ~560 ms). Neither component showed up for the normal sentences.

19

Figure 5. Picture taken from Kutlas and Hillyard (1980). The black line represents the ERP for ‘normal’ sentences. The dotted line is the semantically inappropriate sentences. Note: negative is up.

20

Figure 6. Idealized waveforms for several components of the ERP (taken from Rugg 1995).

21 1.1.4 Source Estimation In section 1.1.1, the inverse problem was introduced. It was noted that any given scalp voltage topography could theoretically be produced by an infinite number of cortical source configurations. That is, the EEG activity we see at the scalp could have any of a number of sources in the brain. Since Helmholz (1853) pointed out the problem it has been understood that, mathematically, there was no way to solve the inverse problem (i.e. no one equation that could take the scalp data and give a specific generator in the brain). With the immense processing power of modern computer, a new technique has been developed, known as source estimation (sometimes called source localization). The strength of source estimation comes from the ability of the computer to model the conductive properties of the brain, dura, skull, etc, and the anatomical configurations of the neurons that create the open fields of the EEG, creating a head volume conductor model. This model is used to “localize” functional generators within the brain. This is known as the forward solution because the models are starting with known conductive properties of the underlying material (brain, skull, neurons etc.) and modifying the input parameters (EEG data) in such a way as to minimize the sum-ofsquares error between the data and model (Slotnick, 2005). A nonlinear model-fitting algorithm is used to iteratively modify the scalp data until a best-fit solution is found. 1.2

Recognition Memory

1.2.1 Prior Occurrence Recognition is the act of perceiving something as previously known (Mandler, 1980). It requires the capacity for a) the identification and b) judgment concerning prior occurrence of an object or event. Mandler (1980) defines recognizing as the process of

22 arriving at a decision about prior occurrence. This study is interested in the underlying neural processes that support recognition memory in humans. Our ability to recognize objects and events as previously encountered is a fundamental to our survival. Case studies of patients with localized damage illustrate this fact in a sad way. In a famous example, Oliver Sacks (1985) reports about a patient with temporal lobe damage who mistakes his wife for a hat! V.S. Ramachandran (2001) reports about patients with damage to the fusiform gyrus who recognize their close relatives but believe that they are imposters. They report a strange feeling that, although the imposters resemble their parents or siblings, they have a keen feeling they are fakes. Such reports are a reminder that our ability to accurately recognize people, objects, and events is a crucial skill for survival. Recognition memory is affected by many factors. Accuracy and performance increases with stimulus frequency, known as repetition priming (see Schacter et al. 1993 for a review, and below). Prior exposure facilitates recognition of items or events. Repetition priming is usually studied in speeded decision tasks (e.g., lexical decision) where priming is indexed by faster reaction times to repeated than unrepeated items, or under conditions that involve degraded viewing, where repetitions increase the probability of likelihood of correct identification. Depth of processing also plays a major role in affecting recognition. Instructing subjects to process items beyond their surface features (e.g., name the item vs. passively viewing it) increases the likelihood of correctly identifying the item in a later recognition memory test (Steward, 1972). Subjects who are asked to judge pictures of a person based

23 on personality score better on recognition tasks than subjects who judge the physical characteristics of pictures of other people (Cain, 2003). 1.2.2 Theoretical Framework Evidence from cognitive, neuropsychological, and neuroimaging studies increasingly indicates that separate memory processes contribute to recognition memory. Most researchers now agree that so called ‘dual process’ theories are needed to account for dissociations found in tests of recognition memory (Yonelinas, 2002). In contrast, ‘single process’ theories, which posit that recognition memory is dependant on only one form of memory, cannot be fully discounted at this time. A single process theorist would argue that dissociations found in the literature reflect quantitative differences in a single memory system, rather than qualitatively different memory systems. A. Imagine that you pass someone on the street and you know that you have met him or her before, but do not recall any specific details about that person (e.g., their name or where you met). Familiarity is usually associated with non-declarative or ‘unconscious’ memory (conceptual implicit memory but not perceptual implicit memory) and is context free, whereas recollection is usually associated with declarative or ‘conscious’ memory (similar but not identical to free recall), and is context dependent (Rugg & Yonelinas, 2003). Various methods have been developed to test a behavioral dissociation of the two forms of memory, and they indicate that familiarity and recollection rely on distinct memory systems. One such method looks at processing speed. It has been found subjects respond faster to judgments based on familiarity (e.g., such as distinguishing between items that were recently studied vs. non-studies items) than to judgments based

24 on recollection (e.g., recollecting specific information about the study event, like when an item was presented in a list of items) (Hintzman, 1998). Recollection depends on relational information about the study event, such as whether a word was initially paired with another context word, or sensory modality or spatial location in which an item was presented. Familiarity does not rely on such information. Jacoby (1991) developed the process-dissociation procedure under the above assumptions to derive quantitative estimates for the contribution of recollection and familiarity. The assumption of this method is that if a subject can recollect a given item, they should be able to determine when or where it was initially studied; familiarity will not reflect such discrimination. Subjects are presented with two classes of items for study. At test, they are either instructed to respond ‘old’ to one class of study stimuli (exclusion task), or they are instructed to respond ‘old’ to all items that have been studied (inclusion task). For the exclusion task, accurate performance reflects the ability of the subject to discriminate between the two classes of study items, which reflect recollection. Recognition based on familiarity is sufficient to discriminate old from new items in the inclusion task. With these assumptions, it is possible to calculate separate values for recollection and familiarity (Jacoby, 1991). Several studies using this technique have shown that recollection and familiarity behave independently (Curran & Hintzman, 1995). In the ‘remember/know’ procedure developed by Tulving (1985), subjects are required to introspect about the basis of their recognition and report whether their recognition stems from remembering (i.e. recollection of specific, episodic information about the study event) or knowing (i.e. having a feeling of experiencing the stimulus in

25 the past in the absence of recollecting any specific information about it). Because subjects are instructed to respond ‘remember’ whenever they recollect a test item, the probability of a ‘remember’ response can be used as an index of recollection, whereas the probability that an item is familiar is equal to the conditional probability that it received a ‘know’ response given it was not recollected (Yonelinas & Jacoby, 1995). The abovementioned behavioral tasks, plus a few others omitted for brevity, show that recognition memory probably relies on different memory processes, but this is stated with a caveat. Although each method is theoretically motivated and has been supported by empirical evidence, it is likely that at least some of the assumptions that each relies on are violated (Yonlias & Jacoby, 1995). Most theorists, when building a theory of recognition, do not rely on data from any one method; instead, they look for convergence of data across a variety of methods. When approached in such a fashion, these methods support the view that recognition memory can be understood as reflecting two distinct memory processes (Rugg, 2003). To better understand the nature of recognition memory, researchers have employed the methods of cognitive neuroscience, which attempts to understand the underlying neural correlates of cognitive processes. Studying the neural correlates of recognition memory would gives researchers a direct way to test ‘dual process’ theories, which predicts that recollection and familiarity should be reflected by different brain regions or neural mechanisms. If familiarity simply reflects a weak form of recollection—as the ‘single process’ theories suggest—then it should have similar neural correlates that differ only quantitatively from those related to recollection (Rugg and Yonelias, 2003).

26 1.2.2.1 Animal Models Distinctions similar to those in the dual process framework have been proposed in the context of research on animal memory. Models based on research with non-human primates propose that the hippocampus is critical for recollection, whereas familiarity is supported by adjacent medial temporal cortex (O’Reilly & Norman 2002; Brown & Aggleton, 2001) The medial temporal lobe (MTL), a region that includes the hippocampal formation and the parahippocampal, enthorhinal and perirhinal cortices, have dissociable mnemonic functions (Rugg & Yonlias 2003). ‘Delayed non-matching to sample’ is profoundly impaired by lesions confined to perirhinal cortex (Meunier et al, 1993). In a seminal study, Brown and Xiang (1998) found that a significant proportion of perirhinal neurons demonstrate lower firing rates for recently experienced than for experimentally novel objects. Evidence such as this has led to the proposal that the perirhinal region plays a crucial role in familiarity-based recognition (Rugg & Yonelias, 2003). Lesions of the hippocampus of non-human primates severely impairs memory for complex associations (e.g. those between a stimulus event and its context), which has led to the suggestion that the hippocampus plays a crucial role in memory for specific episodes, known as recollection here (Eichenbaum, 2000). Research on non-human primates has given researchers a jump start on the neural substrates of human recognition memory, and given them a framework in which to ask whether different regions of the human MTL contribute differently to recollection and familiarity (Rugg & Yonelias, 2003). It should be noted, however, that some models of human memory hold the view that the hippocampus supports both recognition and familiarity.

27 1.2.2.2 Amnesia Much of the neuropsychological evidence for recognition memory in humans comes from the study of patients with global amnesia (i.e. those who, following neurological insult, demonstrate abnormally low scores on test of memory such as the Weschler Memory Scale, but perform normaly on other cognitive functions). Early dual process models predicted these patients should suffer a selective deficit in recollection (Rugg & Yonelias, 2003). This is based on two findings. First, amnesiacs perform normally on implicit memory tasks (Moscovitch, 1993). If implicit memory and familiarity were equivalent, as was assumed, then familiarity should be not be impaired. Secondly, these patients confused recently presented items with frequently presented items, which should be the case if their decisions were based on a strength-like index such as familiarity. Current evidence does not support the view that familiarity is functionally equivalent to implicit memory or that familiarity is preserved in amnesic patients (Rugg & Yonelias, 2003). As expected if recollection is disproportionately affected, amnesiac patients do show greater impairment in relational recognition than item recognition (Downes et al., 2002; Pickering et al., 1989). Disproportionate deficits in free recall compared with recognition have also been reported in some studies, although others have found equal levels of deficits (Volpe, 1986; MacAndrew, 1994). Studies using the processdissociation, ‘remember-know,’ and behavioral testing indicate that memory deficits in amnesiacs are not confined to recollection (Verfaellie 1993; Knowlton 1995). However, the results do suggest that regardless of the fact that both recollection and other memory

28 processes like familiarity and implicit memory are affected, recollection is disrupted to a greater extent (see Yonelinas, 1998; Rugg & Yonelians, 2003, for review). As mentioned above, animal research on recognition memory has led to the proposal that the hippocampus and perirhinal cortex support processes akin to recollection and familiarity, respectively. Relevant evidence for this proposal comes form patients who exhibit damage to the hippocampus due to hypoxia (loss of oxygen). Postmortum studes indicate that such damage is restricted largely to the hippocampus (Cumming, et al., 1984; Zola, et al., 2000). Vargha-Khadem, et al. (1997) found normal or near-normal performance on an item recognition task but very poor recall in patients with hippocampal damage following hypoxia following early childhood, and Mayes et al. (2002) reported similar findings in adult onset selective hippocampal damage. Yet, other studies of post-hypoxic patients using MRI to localize the damage have failed to show a sparing of recognition memory (Verfaelli et al 2000). Several studies looking at relational memory in post-hypoxic patients also turns up similar conflicting evidence. These studies used ‘associative recognition’ tests which requires the subject to discriminate whether a pair of items were originally paired at study. Dissociations between item and associative recognition have been reported in both early onset and adult onset amnesic patients. However, other studies that have contrasted item recognition with associative recognition have found no evidence of relative sparing of item recognition (Stark & Squire, 2003;). In an important study on the matter, Yonelinas, et al. (2002) using coma duration as an index of severity of hypoxia, calculated the covariation between recall, recognition, and severity of hypoxia. The finding was that hypoxic severity predicted the degree to

29 which recollection, but not familiarity, was impaired. Further, estimates from remember/know and other procedures indicated a selective impairment in recollection. These findings give mixed evidence to the proposal that the hippocampus and perirhinal cortex underlie recollection and familiarity. As of late, the reasons for these divergent findings are unclear; several proposals have been put fourth (see Rugg & Yonelinas, 2003). Regardless, the study of amnesia has shown that at least some patients exhibit a selective loss of recollection, which means that recollection can be dissociated from familiarity. The question of whether the hippocampus plays a selective role in recollection awaits further evidence. 1.2.2.3 ERP and Dual Process Models The Event-Related Potentials (ERP) technique has also been used to study the nature of the neural correlates of recognition memory (see section 1.1.2 for a detailed discussion of the ERP technique). Since ERPs lack spatial resolution, they are unable to directly address questions about the neural substrates of recognition, but are still valuable because they can determine whether the neural correlates of recollection and familiarity differ qualitatively—as indexed by ERP effects that differ in scalp distribution rather than magnitude (Rugg & Yonelinas, 2003). These studies usually compare ERPs elicited by recollection and familiarity. Early reports looking at such comparisons offered limited evidence that recollection and familiarity have distinct neural correlates (for a review see Allen & Rugg, 2000), but later studies using a combination of functional, temporal and neuroanatomical grounds have found that the ERP correlates of familiarity and recollection can be dissociated. For example, Rugg et al. (1998) compared ERPs to a

30 task that involved a ‘shallow’ study task in which familiarity process are used with a ‘deep’ study task in which the probability of both familiarity and recollection process would be used. Both classes of recognition judgment elicited an early, frontal effect whereas the deep studied items elicited both the early effect and a late, left-lateralized positive component found over parietal electrode sites. Rugg et al. concluded that these findings show that the frontal effect is a correlate of familiarity, whereas the later parietal effect indexes recollection. Two studies by Curran (2000; 2003) support the findings of Rugg (1998). In a recognition memory test, Curran (2000) had subjects distinguish between ‘old’ (previously seen) and ‘new’ (unseen) words. Some of the words in the test were similar to, but not identical, with the study words (e.g. CAR CARS). These items elicit a high amount of false alarms, and are believed to give rise to strong familiarity in absence recollection (Hintzman, 1994). The ERPs associated with false alarms elicited the early, frontal positivity relative to items correctly judged new, but did not elicit the late parietal effect found in Rugg. On the other hand, ERPs elicited by recognized old items (indexed by correct responses) showed both the early frontal effect and the late parietal effect, leading to the conclusion that familiarity and recollection were dissociated. Similar findings were found in a similar experiment with pictures (Curran & Cleary 2003). 1.2.2.4 fMRI Evidence The ERP studies above suggest that familiarity and recollection depend on separate neural structures that are, at least, partially non-overlapping, but give no direct evidence as there where in the brain these structures might be. Such information can be supplied by fMRI studies. Two studies using the ‘remember/know’ paradigm found

31 increased left lateral parietal activity for old items endorsed as ‘remembered’ rather than ‘know’ (Henson et al., 1999; Eldridge et al., 2000). Together with ERP studies (above), which indicate parietal ERPs correlate with recollection, these fMRI results suggest that lateral parietal region plays a role in this form of memory (Rugg & Yonelinas, 2003). Eldridge et al. (2000) found greater activity in the hippocampus and adjacent MTL cortex for items subjects marked as ‘remembered’ relative to items marked ‘know’ or ‘new.’ In other studies, the level of hippocampal and perirhinal activity elicited by items during study has been a predictor of subsequent recognition memory performance (Davachi et al, 2003). In sum, the data from fMRI studies supports the view that the hippocampus is more active with recollection than when recognition is based on familiarity, but the possibility that the hippocampus plays a role in both recollection and familiarity cannot be ruled out without further evidence (Rugg & Yonelinas 2003). 1.3 Repetition and the Brain Objects are perceived more quickly and easily if you have been previously exposed to them, even in the absence of conscious awareness (Gauthier, 2003). It is thought that exposure of an item “primes” the representation of the object in the brain which makes for easier subsequent processing of the same or similar items (Rugg, 1995). As mentioned above, this phenomena has been studied by speeded decisions tasks in which priming is indexed by faster reaction times (RT) to repeated than unrepeated items, or tasks that require the subject to identify objects under degraded viewing conditions. For instance, Warren and Morton (1982) found that prior presentation of pictures serves to facilitate the recognition of that picture in a test phase up to 45 minutes later.

32 1.3.2 ERP Repetition Effect Almost universally, ERPs evoked by old items elicit more positive waveforms than new items. Based on the theoretical understanding of the ERP, it is assumed that these different patters reflect disparate brain activity. This effect has been studies in indirect memory tasks where performance is influenced by, but is not dependent, on memory for the study item. Indirect memory tasks are usually thought to engage priming effects and, related implicit memory systems. On the other hand, direct memory tasks have also been employed to study the effects of old items on the ERP. In a direct memory tasks make specific reference to a previous learning episode and encourage subjects to reflect on the content of their memory. Direct tasks are thought to engage explicit or conscious memory. There is quite a bit of literature on the modulate of ERPs by the repetition of words and other stimuli in indirect tests (Rugg, 1990, 1998). In a typical paradigm, the subject might be told to respond to occasional ‘target’ (e.g. non-word) items against a background of repeating ‘non-targets’ (e.g. words). In such studies, ERPs evoked by repeated items have been found to be more positive going than those to the first presentation (for a review, see Rugg, 1995). This has been labeled the ERP Repetition Effect. Modulation of at least two major components is normally found: 1) a reduction in the amplitude of the N400 component and 2) an increase in the amplitude of a late positive component usually associated with the P300 (Besson et al 1992; Joyce et al 1998;). Although this finding is consistent across most studies with words, the opposite effect has been found in studies concerned with the repetition of visual objects. For

33 instance, Rugg (1995) found decreased positively to repeated possible geometric objects (i.e., unlike Escher objects). One proposal for the reverse polarity may have to do with involvement of cortical cells with a different orientations from those activated when words or number stimuli are used, while an alternative explanation holds that it reflects reduced neural activity (Penny, 2001). 1.3.3 ERP ‘Old/New’ Effect In direct memory tests, ERPs can be directly related to subjects’ performance. The typical manipulation involves a study-test paradigm, in which items presented in the study phase are intermixed with new items in a test phase, and the subject is asked to respond ‘old’ or ‘new’ to items in the test (the Curran study discussed above is one such example). Numerous studies have described old/new effects in recognition memory and, like the ERP Repetition Effect, most find that correctly classified old items evoke a more positive-going ERPs than to new items (see Yonelinas, 2003 for a review). The ERP old/new effect has traditionally been broken down into three distinct components, probably reflecting different cognitive processes. The first effect, known as FN400, usually occurs bilaterally at frontal electrode sites between 300-500 ms post stimulus and is thought to reflect familiarity (Johnson, 1995; 1998 Friedman and Johnson, 2000). The second component occurs maximally at parietal electrode sites, particularly left, between 500 and 800 ms and is called the parietal effect. This effect is thought to reflect recollection because it is enhanced by items correctly identified as previously studied (Woodruff et al. 2006). Lastly is the late frontal component (LFC), which occurs later in the waveform over the frontal electrode sites, with right hemisphere

34 predominance. The LFC is thought to reflect post-retrieval verification processes (Ally and Budson, 2007). As can be seen, comparing across studies, the effects seen in repetition studies and old/new studies have similar findings (i.e. ERP components). This is most likely because the difference between the repetition paradigm and the old/new is only in the instructions that subjects receive. For instance, in a repetition task, they might be told to identify visually degraded stimuli or read for comprehension, but be told nothing about repetition. In an old/new study-test paradigm, they might be told respond differently to ‘old’ vs. ‘new’ stimuli.

Similar results are probably due, in part, to the same memory

underlying process used by the two approaches (Van Petten & Senkfor, 1996). 1.3.4 Repetition Suppression Contrary to what would be inferred from the above ERP literature, direct measurements of neural activity reveal that when stimuli are repeated, neural activity is usually reduced (for review see Grill-Spector 2006). This effect has been reported at multiple spatial scales—from the level of individual neurons in monkeys to the pooled activity of millions of neurons using fMRI in humans (Naccache & Dehaene 2001). The effect has also been found across multiple temporal scales—from milliseconds to minutes and days—and across many different brain regions, as well as different experimental conditions (Dale et al., 2000). This effect is commonly referred to as repetition suppression or neural priming, the latter because it is thought to use mnemonic processes related to priming and the former because the suppression in neural activity is thought to indicate a “sharpening’ in the neural networks involved in representing objects (Grill-Spector et al., 2006). In other

35 words, new objects are represented by many broadly tuned neurons and, over repetitions, the neurons carrying the least amount of information (presumably, those not necessary for proper representation) drop off. Concurrently, those neurons which are most informative increase their response and become more efficient—hence, the sharpening process (Gauthier, 2003). The ventral temporal cortex, medial temporal cortex and frontal cortex, as well as earlier sensory processing areas, like the primary visual cortices, have shown reductions (see review Kourtzi &Gill-Spector 2005). fMRI responses tend to decrease monotonically (i.e. sequentially) with the number of repetitions, reaching a ceiling effect at about six to eight presentations (Henson et al 2000; Sayres & Grill-Spector 2003). The effect is maximal when there are very few to no intervening stimuli, but it has been observed with tens of interveining stimuli and after multiple days between stimulus presentations (Grill-Sepector et al. 2006). In conjunction with the ERP data on repetition, most of the literature is in concurrence with these findings, but there are a few lone studies that have found the opposite effect. James et al. (2000), is a prime example. Subjects first passively viewed a sequence of 12 objects each appearing for 1 second and repeated ten times. The typical repetitions suppression was found, with decreasing activity as pictures were repeated in the fusiform gyrus (FG), posterior parietal (PP) and frontal lobes (FL). Next, a gradual unmasking paradigm was used in which six of these objects were gradually revealed. James et al. Data compared the effects of priming before and after object identification by taking images before and after subjects felt they could name the object. The masking

36 paradigm was used to slow recognition. Interestingly, repeated objects evoked more activity in the FG and PP than unfamiliar objects prior to identification. 1.3.5 The Current Study It is difficult to reconcile the disparate findings from the three major topics reviewed above. First, it is becoming clear that recognition memory is supported by multiple memory systems, as described by the dual process framework, and it is assume that these different systems are composed by (at least) partially non-overlapping neural mechanisms. Second, recognition performance is enhanced when one have repeatedly encountered an item. In the ERP literature, this enhancement is usually associated with an increase in amplitude of the waveform which is thought to reflect greater activity in the underlying neural generators. On the other hand, fMRI studies give a contradictory picture of what is happening in the brain when items are repeated. In an attempt to reconcile these findings, James et al. (2000) argues that in accordance with their study, the ERP Repetition Effect might reflect the pre-recognition processes, whereas the fMRI data might the reflect post-retrieval period. On the surface, this proposal makes sense, given that the temporal resolution of the ERP technique (on the order of tens of milliseconds) is much greater relative to fMRI (on the order of hundreds of milliseconds to seconds), but further investigation is needed to reconcile these findings. Regardless of the complications of this hypothesis, it does seem likely that ERP data will give a different picture of brain activity than fMRI. fMRI charts relative blood flow to different areas of the brain, time locked to events, which is on the order of seconds. The ERP technique, on the other hand, can detect brain activity in as little at

37 tens of milliseconds which makes it a much better tool for studying rapid, early stages of cognitive processing involved in recognition memory. Using sophisticated analysis software (EMSE, Source Signal Analysis, LLC), we will attempt to use the source estimation technique described in section 1.1.4 to “fit” the electrodynamic brain activity found at the scalp to sources within the brain. Given the high temporal resolution of the ERP technique, we will be able to reveal information about brain activity on time scales that the fMRI is unable to detect. We believe that experimental factors that affect recognition speed should likewise affect the ERP waveforms that are know to be influenced by recognition processes. Further, since these ERP waveforms found on the scalp are generated by the activity of underlying brain tissue, we believe that the same experimental factors that affect recognition will also affect the brain regions involved in recognition processes which will be revealed by a greater signal strength and decreased signal latency at regions implicated in recognition.

38 Chapter 2 Materials and Methods 2.1 Subjects The goal of this study is to investigate the electrical manifestations of the brain as repeated stimuli are presented to the subject. All subjects were undergraduates enrolled at the University of North Carolina Wilmington and were in normal health, right handed, and had normal or corrected-to-normal eyesight. Excluding three subjects who were personal acquaintances of the experimenter, subjects received participation credit for various psychology courses at the university. Subjects were acquired via a sign-up board in the main hall of the psychology building. All subjects signed an informed consent form approved by the UNCW Institutional Review Board. Subjects ranged between 19-42 years, mean = 21.5. Ten female (mean = 21.4 years) and six males (mean = 20.8) participated in this experiment.

39

Figure 7. Pictures of subject taken from the current experiment.

40 2.2 Equipment Data acquisition was made possible by a 64 channel, high resolution analogue to digital ActiveOne® amplifier system and ActiveView® real-time software. Both are developed by BioSemi® (www.biosemi.com). Post-analysis was completed in EMSE Suite—a modular biosignal data analysis program specifically designed to handle and integrate EEG/ERP and MRI data (http://www.sourcesignal.com). Stimuli were presented with Presentation® which is produced by Neurobehavioral Systems® (www.neuro-bs.com). “Active” electrodes are used, which are amplified through the electrode at the source and do not require abrasion of the skin. The ActiveTwo® head cap is a form-fitting cap designed by Dr. Peter Praamstra at the Behavioral Brain Sciences Center, University of Birmingham, United Kingdom. It is made of elastic and is fitted with custom made plastic electrode holders. The electrodes are not integrated into the cap—they must be physically inserted. Sixty four electrode holders are arranged according to the international 10/20 system (refer to figure 9 below). SignaGel® Electrode Gel is a highly conductive past used to form an electrical connection between the electrodes and the subjects scalp.

41

Figure 8. The second generation ActiveTwo® Active Electrode high density EEG system.

42 2.2.1 Electrode Layout Electrode layout follows the standard international 10/20 system (Horman, 1987). Electrodes are labeled according to a standardized scheme that is composed of two parts. The first part is composed of a letter or short string of letters that corresponds to the underlying brain area. For example, Occipital electrodes are designated ‘O,’ and CentralParental areas with ‘CP.’ The second part refers to the electrode placement relative to the midline. Electrodes found lying on the midline end in ‘Z;’ those left of the midline are assigned odd integers and those to the right have positive integers. The values increase the farther one moves away from the midline. Refer to figure 9 for more detail.

43

Figure 9. International 10/20 Electrode Layout (www.biosemi.com) for 64 electrodes.

44 2.3 Stimuli Real world pictures of animals were taken from Google Images. All pictures had their background removed or replaced with a white background in PhotoShop®. Pictures included a broad variety of animals and plants, ranging from coral reef and plant life to bugs to simple animals to apes. In total, 120 different pictures were presented.

45

Figure 10. Example stimuli from the current experiment.

46 2.4 Procedure 2.4.1 Experiment 1 Subjects were sat approximately 36’ in front of a 15” color CRT Apple ViewSmart® monitor. They were shown how to initiate the stimulus presentation program and told to remain seated and to move as little as possible, but to stay relaxed, and to blink normally. Once the program was started, subjects read a set of instructions which told them they would be presented a series of pictures and to simply fixate at the center of the screen. They were asked to concentrate on the pictures and to try to keep their attention directed towards the computer screen.

47

Figure 11. Example Stimulus on screen.

48 Experimental Paradigm The overall architecture of the experiment is a variation on a traditional ERP paradigm (Rugg, 1995). It was composed of a study and test phase, in which subjects encountered a series of pictures. Each picture was presented for one second, with a one second interval. The study phase was composed of three randomized conditions. In condition 1, pictures were presented once; in condition 2, pictures were presented three times; in condition three, pictures were presented ten times. Pictures were randomized across trials and conditions. Each condition contained 30 pictures, for a total 90 different pictures for the study phase, and 420 trials in all. In the test phase, subjects encountered each picture from the study phase once, plus 30 ‘novel’ pictures (i.e. pictures they have not seen before, or new pictures). Pictures were randomly displayed in the test phase for one second, with a one second interval. The 90 original pictures from the study phase were presented plus 30 novel pictures. This paradigm is similar to that used in typical ‘old/new’ ERP experiments except subjects were not instructed to respond to old versus new pictures (for a review of this paradigm see Rugg, 1995,). This omission was made in order to keep the EEG data as “clean” and free from movement artifacts as possible. Considering that clicking involved numerous muscular movements and task monitoring processes would involve numerous extra processes in the brain, it was assumed by the researchers that by omitting overt responses, event related potentials evoked from stimuli would be easier to detect. Several previous studies have shown that ERP waveforms elicited from false alarms are qualitatively different from those elicited by hits (Yonelias, 2003). Therefore,

49 it may be argued, by simply averaging the EEG data for all old pictures, we are conflating false alarm ERP data with ERPs elicited by hits. Although this may be true, the purpose of this experiment was to ascertain whether the brain is responding differently to repeated stimuli. Therefore, if we find significant differences between conditions, it can be argued that hits minus false alarms factored out. Having said that, we collected reaction time data in the exact same paradigm in Experiment 2 (without EEG) in order to test the hypothesis that subjects would respond faster to repeated stimuli.

50

Example of Study:

Seen 10x

Highly Familiar

Seen 3x

Familiar

Seen 1x

Less Familiar

Figure 12. Design of study and test phase of Experiment 1.

Novel

51 EEG/ERP Procedure EEG was recorded from a full array of 64 Ag–AgCl BioSemi (Amsterd, Netherlands) active electrodes embedded in the ActiveTwo electrode cap. EEG activity was amplified with a bandwidth of 0.03-35HZ (3 dB) and digitized at a sampling rate of 512 Hz. Reference in the recording was made to a common average reference (Dien, 1998). The sampling epoch for each trial started 250 ms before onset of stimulus and ended 1800 ms post-stimulus. Raw data files were imported into the EMSE Software Suite (Source Signal Imaging, San Diego, CA) and ERPs were averaged and corrected. Excessive EOG activity was correct for by using the empirical EMSE Ocular Artifact Correction Tool, methodology the investigator learned while attending two separate workshops hosted by Source Signal Imaging, LLC. and CortechSolutions, LLC. The Ocular Artifact Correction Tool allows the investigator to manually distinguish artifact data from clean data. Then, a logarithmic ratio of artifact data versus clean data is produced by EMSE to subtract artifact data where it is detected. Statistical Analysis The statistical analysis of this experiment using the source estimation method discussed in section 1.1.4. Beamformer analysis was performed on averaged files for each subject (four separate conditions for each subject) using four bilateral regions of interest (ROIs) selected from the review of recognition memory and repetition discussed in Chapter 1. The ROIs were the left and right anterior cingulated, left and right medial frontal gyrus, left and right hippocampus, and cuneus and lingual gyrus (primary occipital cortex).

52 Source files for each of the four conditions for each subject were saved. Next, peak detection analysis was performed on each source file. Peaks were defined from previous research using the ‘old/new’ paradigm (Ally & Budson, 2007). A text file was created by the peak detection analysis that contained information about peak amplitude, latency, and area under the curve for each ROI within each condition for each subject. ROIs were broken down into channels so that particular regions could be compared. 2.4.2 Experiment 2 Experiment 2 was identical to Experiment 1, except reaction time data was recorded from subjects instead of EEG. Subjects were given the same instructions from Experiment 1, except instead of being told to passively observe the set of pictures in the test phase, they were instructed to click one of two response buttons on a pad—left click for ‘old’ pictures and right click for ‘new’ pictures. This is the typical “old/new” studytest paradigm (discussed above).

53 Chapter 3 Results Results from experiment 2 are presented first to support the claim that repetition would increase recognition in the visual experiment. Next, visual results are given from the ERP experiment to show the basic ERP old/new effect is present in the data. Finally, statistical analysis from the source estimation technique is provided to support the claim that repetition would increase relative brain activity in selective areas that are involved with the increase in recognition seen in Experiment 2. 3.1 Experiment 2 Figure 13 shows the reaction times for each of the stimulus frequency conditions. Relative to novel pictures, a trend toward faster responding was observed for the pictures that were repeated 3 and 10 times during the study phase of training.

54

1050 1000 950 900 850 800

0

1

2

3

4

5

6

7

8

9

10

Picture Frequency Figure 13: Reaction times as a function of picture repetition frequency during the training phase. A repeated measures ANOVA revealed a significant effect of picture frequency; F(3,33)=12.37, p<.001. Paired comparisons showed that frequencies 0 and 1 differed significantly from 3 and 10 (p<.05).

55

100 90 80 70 60 50

0

1

3

10

Picture Frequency Figure 14. Accuracy as a function of frequency. A repeated measures ANOVA showed that accuracy was affected by picture frequency; F(3,33)=9.14, p<0.001. Paired comparisons showed that 0 (novel) and 1 frequency condition differed significantly from each other (p<0.03) and the 3 and10 frequency conditions both differed from the 0 and 1 conditions (p<0.05).

56 3.2 Visual Results; Experiment 1 3.2.1 Topographic Maps Data from five subjects had to be discarded due to artifacts or technical problems with the equipment. Therefore, only data for 11 subjects were used in the following analysis for Experiment 1. The EMSE analysis software is able to “visualize” the electrical signals found on the scalp using an average brain wireframe taken from fMRI data. As discussed above, direct inferences about underlying neural generators cannot be made from the scalp data. Therefore, statements based strictly on the visual electrical maps would be meaningless. Nevertheless, topographic maps are useful for several reasons. For instance, maps can be displayed over time to see how the electrical activity behaves.

57

Figure 15. Visualization of electrical activity for the Novel condition at early P100 component. As can be seen electrical activity over the Occipital area is dominant. Note: red dots represent electrode locations.

58

Figure 16. Visualization of electrical activity for the late positive ERP old/new effect at 1100 ms post stimulus. Positive signal is seen over mid-to-right hemisphere. Note: red dots represent electrode locations.

59 3.2.2 Grand Averages Grand averaged waveforms are presented from the more typical components in old/new repetition ERP experiments. Preliminary visual inspection of the grand average waveforms reveals findings similar to previous research using the old/new experimental paradigm (see section 1.3.3 for a summary). In short, waveforms elicited by repeated objects shows a positive going trend relative to objects seen for the first time (known here as novel), although this was not consistent across all components. Due to constraints in time, we did not complete the statistical (e.g. area under the curve) analyses on these waveforms. They are included here only to visually demonstrate that our experiment elicited similar trends in the waveform as previous research using similar paradigms. FN400 The FN400 typically occurs at bilateral frontal electrode sites between 300 and 500 ms, and appears to precede controlled attempts by the individual to recollect information. We found at lest three distinct peaks in this time range, especially around the far left side of the frontal lobe (Figure 15). Left hemisphere negativity is positive going relative to repetition, with the largest difference in the 10x condition. Contrary to the typical findings though, we observed little difference on the right hemisphere, with a slight attenuation of the waveform, although no obvious peaks are detectable.

60

Figure 17. Grand average waveforms for all four conditions at electrode cite Af7.

61

Figure 18. Grand average waveforms for all four conditions at electrode cite F5.

62

Figure 19. Grand average waveforms for all four conditions at electrode cite F8.

63

Figure 20. Grand average waveforms for all four conditions at electrode cite F6.

64 Parietal old/new The parietal old/new repetition effect usually occurs around 500 to 800 ms occurring maximally at parietal electrode sites, right hemisphere predominance. Most studies have found a positive going shift in this effect when words were used as stimuli, whereas three studies found the opposite effect for pictures. As can be see below, our results parallel these latter studies in that the parietal waveforms become more negative with repetition.

65

Figure 21. Grand average waveforms for all four conditions at electrode cite P6.

66

Figure 22. Grand average waveforms for all four conditions at electrode cite P8.

67

Figure 23. Grand average waveforms for all four conditions at electrode cite P2.

68 Late old/new Effect The late old/new effect usually occurs around frontal electrodes cites, with right predominance. Visual inspection of the waveforms below shows only small difference between conditions, although a slight positive shift is seen with repetition.

69

Figure 24. Grand average waveforms for all four conditions at electrode cite Af4.

70

Figure 25. Grand average waveforms for all four conditions at electrode cite Af8.

71 3.3 Source Estimation and Repetition Effects 3.3.1 Amplitude Differences The series of figures that appear below present peak amplitudes of five ERP components that previous studies have linked to recognition memory. The bar graphs at the top of each figure show the peak amplitude estimates in each of four brain regions that were generated by the EMSE® beam former source localization software. The regions of interests (ROI) were cuneus cortex and lingual gyrus (which are located in left and right primary visual cortex, respectively), anterior cingulate cortex, hippocampus, and medial frontal gyrus. The panels on the left side represent amplitudes estimates from the left cerebral hemisphere and those on the right side represent ROIs from the right hemisphere. Each ROI was chosen because previous studies suggested that those particular regions play selective roles in visual memory processing (fore review, seee Roland & Gulyas, 1995). ANOVAs and pairwise comparisons were used to statistically evaluate peak amplitude differences between ROIs within each cerebral hemisphere. The bottom panels of each figure show the change in peak amplitudes for each of the picture presentation frequencies relative to the ERP obtained when novel pictures were presented. The data represented in these panels is based on average ERP peak amplitudes across all four ROIs for each cerebral hemisphere.

72

P100 Amplitude

P100 Amplitude

8

8

7

7

6

6

5

5

4

4

3

3

2

2

1

1

0

0

Region (Left Hemisphere)

Region (Right Hemisphere)

P100 Left Hemisphere

P100 Right Hemisphere

1.5

1.5

1.0

1.0

0.5

0.5

0.0 -0.5

1

2

3

4

5

6

7

Picture Frequencey

8

9

10

0.0 -0.5

-1.0

-1.0

-1.5

-1.5

1

2

3

4

5

6

7

8

9

10

Picture Frequencey

Figure 26. ERP peak amplitude of the P100 ERP component was significantly greater in the hippocampus bilaterally than in the other ROIs (p<.05). Picture frequency did not significantly affect peak amplitudes of the P100 ERP component.

73

N155 Amplitude

N155 Amplitude

8

8

6

6

4

4

2

2

0

0

Region (Left Hemisphere)

Region (Right Hemisphere)

N155 Left Hemisphere

N155 Right Hemisphere

1.5

1.5

1.0

1.0

0.5

0.5

0.0 -0.5

1

2

3

4

5

6

7

Picture Frequencey

8

9

10

0.0 -0.5

-1.0

-1.0

-1.5

-1.5

1

2

3

4

5

6

7

8

9

10

Picture Frequencey

Figure 27. The peak amplitude of the N155 ERP component was significantly larger in the hippocampus than in the occipital cortex and medial frontal cortex in the right hemisphere (p<.05) and than the medial frontal cortex in the left hemisphere (p<.05). Picture frequency did not significantly affect the peak amplitude of the N155 ERP component.

74

P300 Amplitude

P300 Amplitude

8

8

7

7

6

6

5

5

4

4

3

3

2

2

1

1

0

0

Region (Left Hemisphere)

Region (Right Hemisphere)

P300 Left Hemisphere

P300 Right Hemisphere

1.5

1.5

1.0

1.0

0.5

0.5

0.0 -0.5

1

2

3

4

5

6

7

Picture Frequencey

8

9

10

0.0 -0.5

-1.0

-1.0

-1.5

-1.5

1

2

3

4

5

6

7

8

9

10

Picture Frequencey

Figure 28. The peak amplitude of the P300 ERP component did not differ across ROIs in the left cerebral hemisphere. In the right hemisphere, however, the peak amplitude of the P300 was significantly larger in the hippocampus than in the occipital cortex or in the medial frontal cortex (p<.05). Picture frequency significantly influenced the P300 peak amplitude in the right hemisphere such that pictures that were presented 3 times generated significantly larger peak amplitudes than novel pictures did (p<.05) and pictures presented 10 times were marginally significantly different from novel pictures (p=0.58).

75

N400 Amplitude

N400 Amplitude

8

8

6

6

4

4

2

2

0

0

Region (Left Hemisphere)

Region (Right Hemisphere)

P400 Left Hemisphere

P100 Right Hemisphere

1.5

1.5

1.0

1.0

0.5

0.5

0.0 -0.5

1

2

3

4

5

6

7

Picture Frequencey

8

9

10

0.0 -0.5

-1.0

-1.0

-1.5

-1.5

1

2

3

4

5

6

7

8

9

10

Picture Frequencey

Figure 29. In the right hemisphere, the N400 was significantly larger in the hippocampus than in the occipital cortex or medial frontal cortex (P<.05). Picture repetition did not significantly affect the N400 ERP component although the mean peak amplitude difference between pictures presented once and novel pictures approached statistical significance (p=.08).

76

Late Frontal Amplitude

Late Frontal Amplitude

8

8

6

6

4

4

2

2

0

0

Region (Left Hemisphere)

Region (Right Hemisphere)

Late Frontal Left Hemisphere

Late Frontal Right Hemisphere

1.5

1.5

1.0

1.0

0.5

0.5

0.0 -0.5

1

2

3

4

5

6

7

Picture Frequencey

8

9

10

0.0 -0.5

-1.0

-1.0

-1.5

-1.5

1

2

3

4

5

6

7

8

9

10

Picture Frequencey

Figure 29. No significant effect of ROI were detected. Pictures presented 1 time differed significantly from novel images (p<.05) and those presented 10 times (p<.05).

77 3.3.3 Peak Latency Differences Figure 29 shows peak latency differences for each ERP component. Left and right hemisphere is presented within graphs. Onset of peak (in milliseconds) is presented as a function of repetition. ANOVAs were used to statistically evaluate peak latency differences between conditions for each ERP component.

78

P100

N155

125

Left Right

120

*

115

P300

200

Left Right

190

320

180

300

110

170

290

105

160

280

100

150

0

1

2

3

4

5

6

7

8

9 10

Picture Frequencey

0

1

2

3

4

5

6

7

8

9 10

Left Right

310

270

* $

0

1

Picture Frequencey

2

N400 Left Right

440

420

1470

410

1460

2

3

4

5

6

7

8

9 10

6

7

8

9 10

Picture Frequencey

1450

*

*

1490 1480

1

5

1500

430

0

4

Late Frontal

450

400

3

Picture Frequencey

0

1

2

3

4

5

6

7

8

9 10

Picture Frequencey

Figure 30. Peak latency is significantly shorter in many of the ERP components. The peak latency for the P100 component was significantly shorter for novel versus 1 condition (p< .05), bilaterally. At N155, latency differences are found in the right hemisphere, novel versus 1 condition. For the P300 component, the latency for the peak at 3 presentations is significantly shorter. We found no difference for the N400 component. The late frontal component had a latency difference for the novel condition versus 10 and condition 1 versus 3.

Left Right

79 Chapter 4 Discussion 4.1

Findings In our experimental paradigm, subjects responded significantly faster and more

accurately to repeated stimuli verses stimuli seen for the first time. This was a graded effect, with those items most familiar producing the fastest response time. The purpose of this study was to determine whether the ERP waveforms and underlying brain regions related to recognition processing would also be affected by the experimental factors that affect recognition. It was proposed that an increase in recognition speed would be correlated with an increase in signal strength and decrease in signal latency for areas of the brain involved in recognition. The data above suggests this hypothesis was confirmed. In the two early effects—P100 and N155—peak amplitudes were found to be largest in the hippocampal RIO, although repetition did not affect the amplitude of these early components. On the other hand, the P300, N400, and LFE all showed repetition effects for peak amplitude, i.e., repetition significantly influenced the amplitude of these components. Picture frequency significantly influenced the P300 peak amplitude in the right hemisphere such that pictures that were presented 3 times generated significantly larger peak amplitudes than novel pictures. Pictures generated significantly larger peak amplitudes in the 1x condition relative to novel condition for the N400 component as well as the LFE. The LFE also showed a difference in peak amplitude for pictures seen ten times versus pictures seen once.

80 Secondly, peak latency was also found to be shorter with repetition, although the findings were not uniform. This effect was found for all conditions. The peak latency for the P100 and N155 components was significantly faster for pictures seen once versus novel pictures. As with peak amplitude, peak latency was significantly different in the 3x condition versus novel for the P300 component. No significant differences in peak latencies were observed in the N400 ERP component. Finally, repetition produced a peak shift for the 10x versus novel condition as well as 3x versus 1x. These data show that repetition influences 1) the amplitude of ERP signals that are generated from underlying brain regions and 2) the timing of these signals, with shorter latency seen for repetition. 4.2

Component Differences The repetition effects are not uniform, as might be expected simply by looking at

the behavioral data. In the behavioral data, subjects responded faster and more accurately as repetition of stimuli increased, in a graded manner. Our statistical findings do not show similar findings with respect to signal amplitude or latency for ERP components. Amplitude was not affected by repetition with early components (P100 or N155). It was with three repetitions for P300, and the LFE showed difference for 1x versus novel and 10x versus 1x conditions. Relative to novel, peak latency was only different for pictures seen once for the two early components, three times for P300 and ten times for LFC. These findings actually make sense when it is taken into consideration that these components probably reflect the processing of information at different latencies in different “memory systems.” What has been shown is that early in the processing stream (P100 and N155) the hippocampus is most activated, but that repetition has no effect on

81 signal strength. As we move down the processing stream, we find that the hippocampus is most activated (right side dominate), and that three repetitions significantly influences signal strength. With N400 the hippocampus in right hemisphere is most active, and 1 repetition influences the signal strength. By 1000ms after the presentation of stimulus, there is no significant difference in the RIO, but repetition affects the signal strength in 1x versus novel and 10x versus 1x conditions. 4.3

Problems Due to artifacts in data, we were forced to discard five data sets. This left us with

only 11 subjects to analyze. Although this is close to the typical number of subjects for ERP studies, it probably affected our ability to detect significant differences in signals due to the source estimation technique we used. In a study that is to follow the current one, we will be doubling our subject pool in order to find out if increasing our data set increases our ability to find significant differences. Most ERP ‘old/new’ studies incorporate behavioral responses into the experimental paradigm. This is done in order to separate ERP responses to hits verses misses and false alarms. We chose not to do this in the current study because we believed that responding would contaminate the EEG data. Since it can be assumed that subjects did correctly recognize stimuli 100% of the time, we probably combined hits, misses, and false alarms into our analysis. While this may have decreased our ability to find significant results, we did find difference between conditions. Therefore, although we may have collapsed hits, misses and false alarms, this did not completely contaminate our findings.

82 In a follow up study to this one, we will first increase the number of subjects by twofold, thereby increasing the chances of finding statistically interesting results. Second, we will be collapsing Experiment 1 with Experiment 2. This way we will be able to separate ERP data for hits, misses and false alarms. 4.4

Future Research The source estimation analysis done in this study is a relatively new approach to

ERP analysis. It is a promising new tool for cognitive neuroscience trying to understand the underlying neural structures that make cognitive processes possible. For the past decade, fMRI has been at the forefront for studying underlying brain regions, but, as noted, this comes at the cost of slow temporal resolutions. We have shown that the source estimation technique can reveal information about the underlying brain regions that fMRI cannot. 4.5

Conclusion Source estimation analysis has revealed that repetition of visual stimuli increases

amplitude signal and amplitude latency from particular regions of interest within the brain. These findings are contrary to fMRI studies that find repetition decreases activity in these regions. We believe that these findings show that ERP source estimation technique can reveal different information about cognitive processes in the brain do to the different temporal scale relative to fMRI.

83

References Allan, K., Robb, W. G., & Rugg, M. D. (2000) The effect of encoding manipulations on neural correlates of episodic retrieval. Neuropsychologia 38(8), 1188-1205. Ally, B. A., & Budson, A. E. (2007) The worth of pictures: using high density event-related potentials to understand the memorial power of pictures and the dynamics of recognition memory. Neuroimage 35(1), 378-395. Ally, B. A., & Budson, A. E. (2007) The worth of pictures: using high density event-related potentials to understand the memorial power of pictures and the dynamics of recognition memory. Neuroimage 35(1), 378-395. Brown, M. W., & Aggleton, J. P. (2001) Recognition memory: what are the roles of the perirhinal cortex and hippocampus? Nature Review Neuroscience 2(1), 51-61. Brown, M. W., & Xiang, J. Z. (1998) Recognition memory: neuronal substrates of the judgement of prior occurrence. Neurobiology 55(2), 149-189. Bruce, V., & Valentine, T. (1985) Identity priming in the recognition of familiar faces. British Journal of Psychology 76 ( Pt 3), 373-383. Cain, K., Oakhill, J. V., & Elbro, C. (2003) The ability to learn new word meanings from context by school-age children with and without language comprehension difficulties. Journal of Child Language 30(3), 681-694. Cipolotti, L. (2001) Long-term retrograde amnesia: the crucial role of the hippocampus Neuropsychologia 39, 151–172. Coles, M. G., Gratton, G., & Donchin, E. (1988) Detecting early communication: using measures of movement-related potentials to illuminate human information processing. Biological Psychology 26(1-3), 69-89. Cummings, J. L. (1984) Amnesia with hippocampal lesions after cardiopulmonary arrest. Neurology 34, 679–681. Curran, T. (2000) Brain potentials of recollection and familiarity. Memory and Cognition 28(6), 923-938. Curran, T. (2000) Brain potentials of recollection and familiarity. Memory and Cognition 28(6), 923-938. Curran, T., & Cleary, A. M. (2003) Using ERPs to dissociate recollection from familiarity in picture recognition. Brain Research and Cognitive Brain Research 15(2), 191-205.

84 Curran, T., & Hintzman, D. L. (1995) Violations of the independence assumption in process dissociation. Journal of Experimental Psychology of Learning and Memory and Cognition 21(3), 531-547. Davachi, L. (2000) Multiple routes to memory: Distinct medial temporal lobe processes build item and source memories Proceeding of the National Academy of Science U. S. A. Dien, J. (1998) Addressing misallocation of variance in principal components analysis of eventrelated potentials Brain Topography 11, 43–55. Donchin, E., Ritter, W., & McCallum, W.C. (1978). Cognitive psychophysiology: The endogenous components of the ERP. In E. Callaway, P. Tueting, & S. Koslow (Eds.),Event-related brain potentials in man. New York: Academic Press. Eichenbaum, H., & Harris, K. (2000) Toying with memory in the hippocampus. Nature Neuroscience 3(3), 205-206. Eldridge, L. L., Knowlton, B. J., Furmanski, C. S., Bookheimer, S. Y., & Engel, S. A. (2000) Remembering episodes: a selective role for the hippocampus during retrieval. National Neuroscience 3(11), 1149-1152. Finnigan, S., Humphreys, M. S., Dennis, S., & Geffen, G. (2002) ERP 'old/new' effects: memory strength and decisional factor(s). Neuropsychologia 40(13), 2288-2304. Fuchs, M., Ford, S Sands, HL Lew (2004) verview of dipole source localization. Physology of Medical Rehabilitaiton of Clinical Neuroscience America. 3, 445-466. Grill-Spector, K., Henson, R., & Martin, A. (2006) Repetition and the brain: neural models of stimulus-specific effects. Trends Cognitive Science 10(1), 14-23. Helmholtz, H. (1853) Ueber eininge Gesetze der Vertheilung elektrischer Strome in korperlichen Leitern, mit Anwendung auf die thierisch-elektrischen Versuche. Annual Physiological. Chemistry 29(3), 222. Henson, R. N., Rugg, M. D., Shallice, T., Josephs, O., & Dolan, R. J. (1999) Recollection and familiarity in recognition memory: an event-related functional magnetic resonance imaging study. Journal of Neuroscience 19(10), 3962-3972. Henson, R. N. A. (1999) Recollection and familiarity in recognition memory: an event-related functional magnetic resonance imaging study. Journal of Neuroscience 19, 3962–3972. Hillyard, L. A., White, H. M., & Abraham, S. (1980) Microcomplement fixation and particle size of chicken lipoproteins. Journal of Lipid Research 21(7), 913-920. Hintzman, D. L., Caulton, D. A., & Levitin, D. J. (1998) Retrieval dynamics in recognition and

85 list discrimination: further evidence of separate processes of familiarity and recall. Memory and Cognition 26(3), 449-462. Holdstock, J. S. (2002) Under what conditions is recognition spared relative to recall after selective hippocampal damage in humans? Hippocampus 12, 341–351. Homan RW Herman J, P. P. (1987) Cerebral location of international 10-20 system electrode placement. Electroencephalography of Clinical Neurophysiology 66(4), 376-382. JACOBY, L. (1991) A process dissociation framework: separating automatic from intentional uses of memory. Journal of Memory and Language 3, 445-456. James, L. E., & Burke, D. M. (2000) Phonological priming effects on word retrieval and tip-ofthe-tongue experiences in young and older adults. Journal of Experimental Psychology of Learning and Memory Cognition 26(6), 1378-1391. Joyce, C. A., Paller, K. A., McIsaac, H. K., & Kutas, M. (1998) Memory changes with normal aging: behavioral and electrophysiological measures. Psychophysiology 35(6), 669-678. Kixmiller, J. S., Verfaellie, M. M., Mather, M. M., & Cermak, L. S. (2000) Role of perceptual and organizational factors in amnesics' recall of the Rey-Osterrieth complex figure: a comparison of three amnesic groups. Journal of Clinical Experimental Neuropsychogy 22(2), 198-207. Knowlton, B. J. a. S., L.R. . (1995) Remembering and knowing: two different expressions of declarative memory. Joural Experimental Psychology Learning Memory Cognition. 51(5). 699-716. Luck, Steve. (Eds). (2004). An Introduction to the Event-Related Potential Technique. Boston: MIT Press. 21, 699–710. MacAndrew, S. B., Jones, G. V., & Mayes, A. R. (1994) No selective deficit in recall in amnesia. Memory 2(3), 241-254. Mandler, G. (1980) Recognizing: the judgment of previous occurrence. Psychology Review 87, 252-271. Manns, J. R. a. S., L.R. . (1999) Impaired recognition memory on the doors and people test after damage limited to the hippocampal region. Hippocampus 9, 495–499. Mayes, A. R. (1999) What are the functional deficits produced by hippocampal and perirhinal cortex lesions? Behavoral Brain Science 22, 460–461. Mayes, A. R. (2002) Relative sparing of item recognition memory in a patient with adult-onset damage limited to the hippocampus. Hippocampus 12, 325–340.

86 Naatanen, R., Gaillard, A. W., & Mantysalo, S. (1978) Early selective-attention effect on evoked potential reinterpreted. Actual Psycholology 42(4), 313-329. Naatanen, R., & Picton, T. (1987) The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure. Psychophysiology 24(4), 375-425. Naccache, L., & Dehaene, S. (2001) The priming method: imaging unconscious repetition priming reveals an abstract representation of number in the parietal lobes. Cereberal Cortex 11(10), 966-974. Nunez, P. L. (1981) A study of origins of the time dependencies of scalp EEG: theoretical basis. IEEE Trans Biomedical Engineering 28(3), 271-280. Pickering, A. D., Mayes, A. R., & Fairbairn, A. F. (1989) Amnesia and memory for modality information. Neuropsychologia 27(10), 1249-1259. Picton, T. W., & Stuss, D. T. (1980) The component structure of the human event-related potentials. Progressive Brain Research 54, 17-48. Pritchard, W. S. (1991) Electroencephalographic effects of cigarette smoking. Psychopharmacology 104(4), 485-490. Ramachandran, V.S., Blakeslee, S. (1999). Phantoms in the brain.: probing the mysteries of the human mind. New York: Harper Perennial. Reed, J. M. a. S., L.R. . (1997) Impaired recognition memory in patients with lesions limited to the hippocampal formation. Behavavoral Neuroscience 111, 667–675. Rempel-Clower, N. L. (1996) Three cases of enduring memory impairment after bilateral damage limited to the hippocampal formation. Journal Neuroscience 16, 5233–5255. Roland, P. E., & Gulyas, B. (1995) Visual memory, visual imagery, and visual recognition of large field patterns by the human brain: functional anatomy by positron emission tomography. Cereberal Cortex 5(1), 79-93. Rugg, M. D., Soardi, M., & Doyle, M. C. (1995) Modulation of event-related potentials by the repetition of drawings of novel objects. Brain Research Cognitive Brain Research 3(1), 17-24. Rugg, M. D., Walla, P., Schloerscheidt, A. M., Fletcher, P. C., Frith, C. D., & Dolan, R. J. (1998) Neural correlates of depth of processing effects on recollection: evidence from brain potentials and positron emission tomography. Experimental Brain Research 123(1-2), 1823.

87 Rugg, M.D. and Allan, K. (2000) Memory retrieval: an electrophysiological perspective. In The New Cognitive Neurosciences (Gazzaniga, M.S., ed.), pp. 805–816, MIT Press. Rugg, M. D., & Yonelinas, A. P. (2003) Human recognition memory: a cognitive neuroscience perspective. Trends in Cognitive Science 7(7), 313-319. Sacks, Oliver., (1985) The man who mistook his wife for a hat: and other clinical tales. Simon & Schuster, Inc. Schacter, D. L., Chiu, C. Y., & Ochsner, K. N. (1993) Implicit memory: a selective review. Annual Review of Neuroscience 16, 159-182. Slotnick, S. D., & Dodson, C. S. (2005) Support for a continuous (single-process) model of recognition memory and source memory. Memory and Cognition 33(1), 151-170. Slotnick, S. D., & Dodson, C. S. (2005) Support for a continuous (single-process) model of recognition memory and source memory. Memory and Cognition 33(1), 151-170. Stark, C. E. (2002) Recognition memory for single items and for associations is similarly impaired following damage to the hippocampal region. Learning and Memory 9, 238–242. Stark, C. E., Squire, L.R. (2003) Hippocampal damage equally impairs memory for single items and memory for conjunctions. Hippocampus 13(281–292). Stark, C. E., & Squire, L. R. (2003) Hippocampal damage equally impairs memory for single items and memory for conjunctions. Hippocampus 13(2), 281-292. Steward, W. E., 2nd, Gosser, L. B., & Lockart, R. Z., Jr. (1972) The effect of priming with interferon on interferon production by two lines of L cells. J Gen Virol 15(1), 85-87. Tulving, E. (1985) How many memory systems are there? American Psychologist 40(4), 385398. Vargha-Khadem, F., Gadian, D. G., Watkins, K. E., Connelly, A., Van Paesschen, W., & Mishkin, M. (1997) Differential effects of early hippocampal pathology on episodic and semantic memory. Science 277(5324), 376-380. Verfaellie, M., Treadwell, J.R. . (1993) Status of recognition memory in amnesia. . Neuropsychology 7, 5–13. Verfaellie, M. M. (2000) Acquisition of novel semantic information in amnesia: effects of lesion location. Neuropsychologia 38, 484–492.

88 Warren, C., & Morton, J. (1982) The effects of priming on picture recognition. British Journal of Psychology 73(Pt 1), 117-129. Warren, C. M., J. (1982) The effects of priming on picture recognition. British Journal of Psychology 73(1), 117-129. Yonelinas, A. P. (1998) Recollection and familiarity deficits in amnesia: convergence of remember-know, process dissociation, and receiver operating characteristic data. Neuropsychology 12(323–339). Yonelinas, A. P. (2002) Effects of extensive temporal lobe damage or mild hypoxia on recollection and familiarity. Nature Neuroscience 5, 1236–1241. Yonelinas, A. P., & Jacoby, L. L. (1995) Dissociating automatic and controlled processes in a memory-search task: beyond implicit memory. Psychological Research 57(3-4), 156-165. Zola, S. M. (2000) Impaired recognition memory in monkeys after damage limited to the hippocampal region. Journal of Neuroscience 20, 451–463. Zola, S. M., Squire, L. R., Teng, E., Stefanacci, L., Buffalo, E. A., & Clark, R. E. (2000) Impaired recognition memory in monkeys after damage limited to the hippocampal region. Journal of Neuroscience 20(1), 451-463. Zola-Morgan. (1986) Human amnesia and themedial temporal region: enduring memory impairment following a bilateral lesion limited to field CA1 of the hippocampus. Journal of Neuroscience 6, 2950–2967. Zani, A., Proverbio, A. (Eds.). (2002). The Cognitive Electrophysiology of Mind and Brain. New York: Academic Press.

STIMULUS REPETITION EFFECTS ON PICTURE ...

Raw EEG data from individual subjects reflects “spontaneous” fluctuations of ... independently, but whose open fields summated at the time recorded (Rugg, ...

4MB Sizes 0 Downloads 234 Views

Recommend Documents

Masked repetition and phonological priming in picture ...
picture naming was facilitated by the prior masked visual ... Due to the hypothetical time course of information flow in this model .... 2 factorial design. Prime–target pairs were coun- terbalanced across the priming conditions across two groups o

Additive and Interactive Effects of Stimulus ... - Semantic Scholar
and Social Sciences, Swinburne University of Technology, Melbourne,. Victoria, Australia ... network gives a “word” response to a nonword). More generally, ..... adaptation would be a modeling enterprise in its own right and is clearly beyond ...

Effects of Fiscal Stimulus in Structural Models
Jul 8, 2011 - International Monetary Fund, the European Commission, the OECD and ... the state of public finances after the 2008 crisis. 2 ...... utility function.

Subjective organization and effects of repetition in ... - Alice Kim, PhD
University of Toronto, Toronto, Canada ... ing (FRL) experiments at Toronto. One of the major ..... Society, Bryn Mawr College, August 29-31, 1963. MILLER, G. A. ...

Additive and Interactive Effects of Stimulus ... - Semantic Scholar
manipulation of a single parameter (i.e., feature-to-letter activa- tion). ..... ences between the joint effects of stimulus quality and word frequency in reading aloud ...

Additive and Interactive Effects of Stimulus ... - Semantic Scholar
S. O'Malley and D. Besner (2008) showed that additive effects of stimulus degradation and word frequency in reading aloud occur in the presence of nonwords but not in pure word lists. They argued that this dissociation presents a major challenge to i

Influence of stimulus amplitude on unintended ...
on the degree to which the actor visually tracks the stimulus. ... Human Movement Science xxx (2011) xxx–xxx ..... A computer recorded the movement trajectories of the forearm and controlled ..... Post, A. A., Peper, C. E., & Beek, P. J. (2000).

Fuzzy Picture Processing: Effects of Size ... - Semantic Scholar
Viconi for her help in data acquisition, and all participants for having taken part in the studies. ... comparing the effects of size reduction to the effects of low-pass ... Pictures were presented using E-Prime software (Schneider, ... Data analysi

Fuzzy Picture Processing: Effects of Size ... - Semantic Scholar
16.22° vertical degrees of visual angle) to very small (2.68°. 2.05°) sizes. Moreover, additional perceptual degradation con- ditions were obtained by applying a ...

When does size not matter? Effects of stimulus ... - Wiley Online Library
Similarly, recent studies showed degrees of emotional modulation of autonomic responses and ... distance and retinal size are strictly related (Loftus & Harley,.

Repetition Maximization based Texture Rectification
images is an essential first step for many computer graph- ics and computer vision ... matrix based rectification [ZGLM10] can be very effective, most of our target ...

Repetition Maximization based Texture Rectification
Figure 1: The distorted texture (top) is automatically un- warped (bottom) using .... however, deals in world-space distorting and not with cam- era distortions as is ...

Gauging the Effects of Fiscal Stimulus Packages in the ...
Sep 10, 2012 - analysing the economic effects of the fiscal stimulus packages. ... Based on a growth accounting exercise, as detailed in Coenen, Straub and Trabandt ... variables, including a small number of foreign variables, which, in the ... inter

Productivity Effects on Mexican Manufacturing ...
greater degree of openness of the Mexican economy that supposedly enhanced productivity: first ... effects of international technology diffusion on productivity. 1.

Environmental Effects on Oxygen Isotope ... - Plant Physiology
evidence supports the theory of the Péclet effect. (Barbour et al., 2000b, 2004; Barbour and Farquhar,. 2003). Other models have been developed to explain a.

Network Effects on Worker Productivity
May 19, 2016 - decisions faced by personnel managers, e.g. how training policies should be optimally designed. .... 5, we describe how we define and construct our co-worker networks. Section 6 is ...... on the same floor of the building.

University Effects on Regional Innovation
Apr 2, 2012 - systems, it is often suggested that stimulating academic research and .... characteristics of regional technological infrastructure (business ... used as controls, and to assume that they are sufficient to exclude ...... Small Business.

Entrenchment & Memory Development Effects on ... -
needs to divide the phoneme sequence first into morphemes and words and, at a higher level, into noun phrases in which gender agreement must be ...

clay effects on porosity and resistivity
problem is, however, especially bad in the interpretation of resistivity data, and also affects the porosity logs. ... resistivity values, but because such data effects the final calculated STOOIP for a given formation. Even small .... effective poro

anthropogenic effects on population genetics of ... - BioOne
6E-mail: [email protected] ... domesticated status of the host plant on genetic differentiation in the bean beetle Acanthoscelides obvelatus.

Supersymmetric electro-weak effects on
m,/Mw sin c~, and so g2 w ,~ mu 2, i.e. ga w and g2 fv areof the comparable size. ..... for an array of sneutrino masses (from 11 GeV to. 91 GeV) and for A = 3 - x/3 ...

PhD scholarship opportunities on climate change effects on Australian ...
PhD scholarship opportunities on climate change effects on Australian marine ... is available to support several PhD projects but candidates need to obtain a.

EFFECTS OF SURFACE CATALYTICITY ON ...
The risk involved, due to an inadequate knowledge of real gas effects, ... the heat shield surface, increase the overall heat flux up to about two times, or more, higher than ..... using data from wind tunnel and free flight experimental analyses.

On letter frequency effects
Aug 19, 2011 - be stronger for token frequency counts compared with both type frequency and frequency rank, stronger for frequency .... Connectionist models of letter and word processing. (such as the ... Stimuli and design ..... Database.