Scott H. Plantz, MD, FAAEM

• Associate Clinical Professor of Emergency Services
• Rosalind Franklin University of Medicine and Science
• Chicago Medical School
• Chicago, Illinois

Spatial cognitive maps in animals: New hypotheses on their structure and neural mechanisms blood pressure 55 years age purchase digoxin 0.25mg otc. Remote spatial mem ory in an amnesic person with extensive bilateral hippo campal lesions blood pressure medication olmetec purchase digoxin on line. The human retrosplenial cortex and thalamus code head direction in a global reference frame arteria femural purchase generic digoxin online. Compli mentary roles of the hippocampus and retrosplenial cor tex in behavioral context discrimination arteria jackson order discount digoxin line. Place field repetition and purely local remapping in a multicompartment environment pulse pressure congestive heart failure purchase digoxin 0.25mg on-line. Transient and enduring spatial representations under disorientation and self rotation hypertension 180100 order 0.25mg digoxin mastercard. Dissociable retrosplenial and hippocampal contributions to successful formation of survey representations. Hippocampal replay captures the unique topological structure of a novel environment. This is because quantity is central to human rationality, and numeri cal concepts are the bedrock of all human measurement- number "measures all measurables," as Locke says. Whether measur ing sets, time, distance, size, weight, or value, humans primarily use numerical scales to formalize and unitize quan tities. Numbers are abstract representations that describe incremental changes in object quantity and that can be logi cally evaluated and transformed. Simple logical operations on numbers, such as comparison and arithmetic, are the building blocks of human mathematics. Substantial evidence indicates that numerical value can be represented without language, in an analog format, and is cognitively manipu lated using nonlinguistic logical operations. This primitive arithmetic exists in modern humans in a psychological and neural format similar to other species. However, human cul tures symbolically formalize numerical relations that have a unique impact on human cognition, behav ior, and brain activity compared to other species. We present research from the field of numerical cognition across multiple levels of analysis to understand the mutual interactions between its origins and purpose and its computations and biology. Developmental Basis Studies on human newborns and preverbal infants sug gest that domain knowledge about numerical relations establishes the foundation of numerical development in humans. Neonates, just hours after birth, can dis criminate the numerical values of sets nonverbally with crude acuity. Izard, Sann, Spelke, and Streri (2009) showed that newborn infants look longer at visual arrays that numerically match the number of sounds they hear in an auditory sequence compared to numerically dif ferent visual arrays. The study showed that newborn infants represent numerical value at an abstract percep tual level across modalities. Several studies of older infants have produced results that show the early repre sentation of number (Barth et al. The implication is that experience expectant cognitive processes detect quantitative variation in sets and events at birth. These studies raise questions about how infants, and humans more generally, disentangle numerical repre sentations from other correlated information in the environment. There are natural correlations between quantitative dimensions in the environment (Cantrell & Smith, 2013; Ferrigno et al. Infants are sensitive to quantitative dimensions beyond numerical value, includ ing surface area, duration, and density (Clearfield & Mix, 2001; Cordes & Brannon, 2008; Lourenco & Longo, 2010). These dimensions also provide valuable quantitative the origins and organization of numerical concepts are studied integratively at multiple levels of analysis. This is important because there are interacting constraints on the mechanisms the brain can implement. This approach is necessary because it accounts for dif ferent pressures- evolutionary and developmental, neural and functional, environmental, and algorithmic-that limit the mechanisms the brain can or will implement. The field of numerical cognition not only investigates the underlying domain representa tions but also examines the ways those representations arise from the dynamic interaction between genetic con straints and environmental input. In this review we dis cuss the dif ferent levels of analy sis at which numerical cognition is understood. We show comparisons of 817 information about sets and events and they are often cor related with number. For example, a set of six figs often (but not always) has a greater number, cumulative sur face area, and volume than a set of three figs. Some have argued that infants are initially "one bit" and only repre sent a general magnitude value across different dimen sions including number, area, and duration (Cantrell & Smith, 2013; Walsh, 2003). Infants are thought to learn to disentangle quantitative dimensions from correlation patterns in the environment. However, how an infant would ever disentangle correlated dimensions without first making some prediction about or interpretation of the underlying components is unclear. For example, in order for infants to detect breeches of correlated struc ture among dimensions they would have to know that multiple different quantities exist. Thus, it is as yet unclear what algorithm or process might permit infants to develop representations of number from "one bit. A base quantity of one provides a foundation for calcu lating all integers by adding one to one, and so on, up to any size. Another proposal for innate knowledge is that the algebraic properties of neural codes for numer osity inherently represent arithmetic relations (see Hannagan et al. Models of number coding are described further in the section on algorithmic mod els; however, the point here is that some theoretical proposals about the nature of innate numerical knowl edge require only simple psychological constraints. Evolutionary Basis the extensive literature on numerical abilities in non human animal species converges with developmental data from human infants in an evolutionary interpreta tion of the origins of numerical cognition. In several studies with newborn chicks, chicks raised in controlled environ ments imprinted on a set of objects and followed that set as their "mother". Once those chicks imprinted on a set, the experimenters tested them in trials with novel "mother" sets that varied in numerosity. The results showed that the chicks established their imprinting response on numerosity-they were more likely to follow sets with similar numerical values to their original "mother. Birds trained on quantity dis crimination tasks in the lab show sensitivity to numerical value when tested with stimuli that are controlled for alternative cues like surface area. Quantitative abilities in young animals suggest that a core function of animal brains is to compare amounts. Indeed, many species compute amounts of various types- even worms are sensitive to differences in ion concentration (Sambongi et al. The simple logic of quantity comparison is likely widespread across dif ferent ner vous systems. Primates have sophisticated numerical abilities and are likely to share homologous cognitive, neural, and developmental processes with humans. The ability to make numerical choices develops rapidly and sponta neously in nonhuman primates. Infant monkeys are able to make reliable quantitative choices within one year of life (Ferrigno, Hughes, & Cantlon, 2016). Numerical development is thus a primitive and rapidly emerging aspect of all primate cognition. Primates have been shown to engage in a range of logical operations with numerical values. Behavioral research with lemurs, monkeys, and apes shows they possess logical capacities for comparison, increment ing, ordination, proportion, and addition and subtrac tion with quantities (Beran, Parrish, & Evans, 2015; Cantlon et al. When monkeys com pare visual arrays of dots to determine the smaller quantity, their per for mance closely resembles that of human subjects who are prevented from counting. Some of the more complex arithmetic abilities of primates are proportional reasoning, addition, and subtraction. Monkeys can compare the relative lengths of two pairs of lines to determine whether the propor tion relation between pairs is similar (Vallentin & Nie der, 2008). Apes and monkeys can predict the arithmetic outcome of sets combined behind an occluder. For example, if 6 items are covered by an occluder, then 3 more items are added behind the occluder, monkeys will guess that there are 9 items behind the occluder (versus 3, 6, or 12). Monkeys also track the relative val ues of sets during oneby one set construction, showing 818 Concepts and Core Domains an ability to represent countlike incremental changes in numerical value (Cantlon et al. Finally, mon keys can make metacognitive judgments about their accuracy during numerosity tasks, and thus their numerical processes are available to internally monitor (Beran, Smith, Redford, & Washburn, 2006). These capacities in nonhuman primates suggest that several logical tools for quantitative cognition emerged many millions of years ago in the human lineage. For example, baboons make collective troop movements by estimating the number of individual animals in a subgroup that took each of a few possible paths and choosing the greatest number (Strandburg Peshkin et al. As the difference (Weber fraction) between the number of baboons in each subgroup increases, animals are more successful at choosing the larger group. These findings are evidence that numerical comparison is com puted naturally by wild primates. Numerosity abilities have been observed in so many species that it would be newswor thy to discover a species that lacked it. Fish have been shown to use numerical comparisons during schooling and collective behav iors. Even insects and other invertebrates are suspected to use numerical representations in their natural behav iors (Chittka & Geiger, 1995; Gallistel, 1990; Wittlinger, Wehner, & Wolf, 2006). However, it is currently unclear whether true numerical reasoning is involved in many of these other cases versus rate, duration, mass, or den sity perception. A recent study that directly compared spontaneous numerical reasoning in human adults from different cultures, children, and monkeys reported significant qualitative similarities in numerosity per ception between groups, but such direct comparisons have not been conducted with nonprimate animals (Ferrigno et al. Evolutionary simulations of numerical cog nition show a plausible route to numerosity representa tion through natu ral selection. Hope, Stoianov, and Zorzi (2010) used artificial life simulations built on the hypothesis that quantity comparison originated from foraging adaptations to maximize food intake. The model shows that numerical sensitivity could plausi bly emerge by genetic selection for foraging efficiency over evolution. Numerical reasoning could also underlie aspects of social behav ior (McComb, Packer, & Pusey, 1994; Wil son, Hauser, & Wrangham, 2001). Social playback experiments show that animals such as chimpanzees and lions use "number of calls" as a cue for deciding intergroup confrontations. When lions were played a small number of foreign lion calls from a hidden speaker, they were more likely to confront the source than if played a large number of calls (McComb, Packer, & Pusey, 1994). Both the social and foraging functions of numerical reasoning show that objectbased, crossmo dal quantity judgments in natural environments likely shaped the design of numerical mechanisms. C, Dehaene and Changeux (1993) modeled numerical representation in a neural network. Visual objects in an array stimulus are first normalized to a location and sizeindependent code. Numerosity detectors are connected to summation activation, and neural activity is tuned to numerosity in an on center, off surround pattern. An open question is whether and how summation neurons and tuning neu rons work together to represent numerical value (Piazza & Izard, 2009). One possibility is that summation neu rons accumulate entities to compute a set representa tion, and tuning neurons place those sums within the relative context of a number line. Human neuroimaging studies indicate that the representation of numerosity also occurs spontaneously early in child development, in a parallel network of neu ral regions (figure 70. Such findings suggest that evolutionarily primitive and early developing properties of frontoparietal circuits are responsible for the emer gence of number coding neurons. Human adults and children show neural tuning to numerosity in functionally overlapping regions of intra parietal cortex (Kersey & Cantlon, 2017; Piazza et al. Similarly, nearinfrared spectroscopy and electroencephalogra phy studies of infants in the first year show numerical sensitivity in the right parietal cortex (Edwards, Wag ner, Simon, & Hyde, 2016; Izard et al. Observations of summation neurons in humans could require more granular data than those currently available. The prefrontal and parietal cortices are regions that process stimuli at a high level of perceptual and motor abstraction in primates (Nieder, 2016). Numerical repre sentation requires abstraction across object and event features, including space, time, perspective, and modal ity, to represent a "set. The parietofrontal network observed in numerical processing in humans and monkeys is known to meet these demands because those regions take inputs from multiple sensory and perceptual regions, have large spatial and temporal receptive windows, and provide abstract outputs to premotor structures (Cavada & GoldmanRakic, 1989; Hasson, Yang, Vallines, Heeger, & Rubin, 2008). Parietal regions also show biases toward topographic representation, and numerosities appear to be topographically mapped there, which could be criti cally related to the ordinality of number (Harvey, Klein, Petridou, & Dumoulin, 2013). Intracranial electrocorticography recordings from intraparietal cortex in three epilepsy patients showed numberrelated neural activity during natural numeri cal reasoning over the course of 7 to 10 days (Dastjerdi et al. The subjects were implanted with chronic intracranial electrodes covering lateral parietal cortex while being continuously monitored by video record ing. Each electrode captured a signal from a popula tion of around 500,000 parietal neurons. Physiological interventions in monkeys suggest that posterior parietal cortex plays a causal role in numeri cal representation. Brief periods of pharmacological inactivation to posterior parietal cortex (area 5) with muscimol caused monkeys to underestimate the num ber of items in a sequence of movements (Sawamura, Shima, & Tanji, 2010). The underestimation was not caused by impairment in motor control because the monkeys were able to perform correct movement types in response to an auditory tone-they only failed to produce the correct number of movements. Human neuropsychological data also show that focal lesions to posterior parietal cortex cause number specific deficits (Dehaene & Cohen, 1997). Together, those data indi cate that neural signatures of numerical processing from posterior parietal regions are not simply correla tional but causal. Neurons within the nidopallium cau dolateral fire with a pattern similar to neural tuning responses in primates; however, the underlying neural anatomy is distinct. These findings from birds show that there are at least two similar yet independently evolved neural implementations of numerical representation in the animal kingdom (Nieder, 2016).

Brain state refers to the internal state of the individual blood pressure goes up and down order digoxin on line amex, which we include as a kind of situational cue hypertension causes buy digoxin cheap online, such as mood (Bower arrhythmia heart disease generic digoxin 0.25mg with visa, 1981; Eich blood pressure 9050 buy discount digoxin 0.25 mg online, 1995) hypertension nursing interventions buy discount digoxin online, hormonal state (McGaugh hypertension yeast infection order 0.25mg digoxin visa, 1989), or feelings associated with the administration of drugs (Overton, 1964). Whether external situational cues, such as the normative rules surrounding an event, and internal situational cues, such as the brain state, have qualitatively dif ferent influences on contextual representations remains an open question. Temporal cues Every thing we do occurs at some time, and it is possible to remember that different events that occurred in the presence of similar spatial or situational cues occurred at different times. Time of day can serve as an important mnemonic cue in spatial memory tasks (Boulos & Logothetis, 1990). Time- of- day effects are also observed in contextual fear- conditioning experiments that interrogate episodic memory, in which animals learn to fear a spatial context in which shock was previously experienced. Rodents display their strongest context- dependent fear response during their inactive phase (the light period; Chaudhury & Colwell, 2002). The second kind of temporal cue is the relative sequence in which learning takes place. Events experienced closer together in time are more similar than events experienced further apart. As a result, if a person experiences an event and her memory is later assessed, the ability to recall that event will decrease as the time between learning and retrieval increases (Rubin & Wenzel, 1996). Similarly, items encountered in close temporal proximity are more likely to be recalled sequentially than items encountered further apart (Howard & Kahana, 2002). This brief taxonomy of context- defining cues suggests that context is characterized by factors external to the agent, including the set of environmental cues that define a place or the situation that characterizes an event, and the internal factors. The cinema provides an apt metaphor for summarizing these context- defining cues: it contains multiple movie theaters (spatial cues) playing different movies (situational cues) at different times (temporal context) (figure 19. For context to be a useful scientific construct, there must be factors that differentiate contexts from other types of mnemonic cues. We suggest three important properties that limit the appropriate application of the term context. First, for the brain to form contextual representations from statistical cue regularities, the cues that characterize context must be reliably present over time, or stable (Biegler & Morris, 1993; Robin, 2018; Stark, Reagh, Yassa, & Stark, 2017). For instance, the location of seats that define a movie theater context must not change often for the seat locations to form an integral part of that context. However, if they are first preexposed to the context, the shock elicits a fear response when the animal is subsequently returned to the conditioned context. Contextual conditioning thus only occurs if animals have an opportunity to learn the reliability of contextual cues through prolonged or repetitive exposure, indicating that the experience of cue stability is critical for the formation of contextual representations that organize memory. The longer rodents experienced a context prior to fear conditioning, the more likely they were to show contextual conditioning (% freezing). When participants recalled locations of landmarks in a city, their recall patterns showed evidence of hierarchical clustering into multiple smaller local contexts. Landmarks were drawn closer together on a map when recalled as being in similar local contexts (Within) than in different local contexts (Between) (Hirtle & Jonides, 1985). Second, just as eating popcorn does not define being in a cinema (one can also eat popcorn at home), contexts are not defined by any single discrete cue (Robin, 2018). In other words, contexts are not the same as cues that serve as discrete signals for other events. Unlike contexts, increased time spent with a discrete cue does not alter conditioning to that cue (Fanselow, 1990). The context of your local movie theater could be recalled as such independent of whether you have popcorn, or are seeing a horror or a comedy film, or have consumed caffeine beforehand. This corollary suggests that context is not simply the set of cues associated with a particular event but rather a holistic representation of those cues. Therefore, context is a neural construct, rather than something that exists in the world (Anderson, Hayman, Chakraborty, & Jeffery, 2003). As an illustration of this point, suppose the locations of the seats in your local movie theater are moved in your absence. The answer to this question is not knowable a priori, but you could easily answer this question about your own memory. Third, because contexts are not defined by any one discrete cue, different context- defining cues must have a reliable organization that allows them to be unified in a contextual representation. A common cue organization used by the brain to represent contexts is a hierarchy (Jeffery, Anderson, Hayman, & Chakraborty, 2004; Pearce & Bouton, 2001). There is an extensive literature demonstrating that the spatial environment is encoded as multiple hierarchically organized contexts, varying in spatial scale, instead of a single environmental context, and per for mance on memory tasks is influenced by this hierarchical structure (Han & Becker, 2014; Hirtle & Jonides, 1985; Holding, 1994; Marchette, Ryan, & Epstein, 2017; McNamara, 1986; McNamara, Hardy, & Hirtle, 1989; Montello & Pick, 1993; Wiener & Julian and Doeller: Context in Spatial and Episodic Memory 219 Mallot, 2003; figure 19. Purchasing movie tickets or purchasing movie snacks are both subordinate to the larger class of transactional situational contexts, and the relative sequence of events can be organized over minutes or days. Beyond hierarchical arrangements, the set of possible relational structures between cues necessary for such cues to be associated in a contextual representation is unknown. An important area for future research is the extent to which different context- defining cues, matched in terms of their behavioral relevance-not just in an experimental situation but also over the lifetime of an individual or evolution- are incorporated into contextual representations. The Hippocampal Basis of Contextual Memory There is consensus that the hippocampus in the mammalian medial temporal lobe plays a crucial role in spatial and episodic memory, and neurobiological studies of contextual processing have focused on this brain area (for reviews, see Maren, Phan, & Liberzon, 2013; Myers & Gluck, 1994; Ranganath, 2010; Rudy, 2009; Rugg & Vilberg, 2013; Smith & Mizumori, 2006; Winocur & Olds, 1978). In the 1970s, Hirsch (1974) first explicitly proposed that the hippocampus mediates the retrieval of information in response to contextual cues that refer to the retrieved information. Since then, a wide variety of studies in both human and nonhuman animals have reinforced the importance of the hippocampus for context- dependent memory. Consistent with these neuroimaging findings, lesion studies have shown that the hippocampus is necessary for maintaining context- dependent memories (Anagnostaras, Gale, & Fanselow, 2001; Maren, 2001). When rodents are conditioned in one spatial context, for instance, they typically show a reduction of conditioned responses when tested in a new context, but animals with hippocampal damage continue to respond as if they failed to notice the spatial context change (Bachevalier, Nemanic, & Alvarado, 2015; Butterly, Petroccione, & Smith, 2012; Corcoran & Maren, 2001; Honey & Good, 1993; Penick & Solomom, 1991). Finally, hippocampal lesions impair the ability to recall the biological time of day at which an event occurred (Cole et al. Based on this survey of context- defining cues and their boundary conditions, we offer the following inclusive definition of context: Context is a holistic representation of the internal and external (stable, nondiscrete, and reliably organized) cues that predict par ticular behavioral or mnemonic outputs. This definition unifies the contextual cues by placing emphasis on the adaptive function of contextual representations, rather than on any one specific cue type (Mizumori, 2013; Stachenfeld, Botvinick, & Gershman, 2017). Note that although this definition runs the risk of circularity, we have proposed three boundary conditions that limit the correct application of the context construct- stability, nondiscreteness, and reliable organization-and immunize against circularity. Insofar as the role of context is concerned, this definition is consistent with theories of memory that do not place particular importance on any one contextual cue type but rather focus on the function of contextual representations (Eichenbaum, 1993, 1996; Howard & Kahana, 2002; Mensink & Raaijmakers, 1988; Schacter, 2012; Schacter, Addis, & Buckner, 2007; Ranganath, 2010). For instance, when recalling previously read scenarios, participants spontaneously generate spatial contexts for the scenarios, even when the scenarios did not include any spatial cues (Robin, Buchsbaum, & Moscovitch, 2018; see also Hebscher, Levine, & Gilboa, 2017). However, as eluded to above, the situational and the temporal context can also strongly influence memory if they are behaviorally relevant. B, Contextual memory is indexed by hippocampal remapping, in which all simultaneously recorded neurons alter their firing patterns across contexts (Alme et al. Hippocampal neurons represented locations in two different situational contexts, one relative to a moving platform (left) and another relative to the stable room (right; Keleman & Fenton, 2010); (3) Temporal cues. When rodents explored two chambers containing objects in different positions associated with different valences, hierarchical cue structure was reflected in hippocampal population activity patterns (McKenzie et al. Thus, the hippocampus is necessary for the retrieval of memories associated with contexts characterized by the full range of context- defining cues. Neuroimaging studies in humans likewise support the idea that the hippocampus represents a map of local context (Epstein, Patai, Julian, & Spiers, 2017). Beyond distinguishing between locations within a context, however, the hippocampus also stores multiple maps that allow it to represent multiple contexts (Bostock, Muller, & Kubie, 1991; Muller & Kubie, 1987). During remapping, when an animal changes contexts, all simultaneously recorded neurons shift their relative place fields to new locations or stop firing altogether, quickly forming a new map-like representation (Bostock, Muller, & Kubie, 1991; Save, Nerad, & Poucet, 2000). If remapping mediates contextual memory, then remapping should occur between contexts defined by all contextual cue types and should be constrained by the same factors that limit when cues do not define contexts. Interestingly, rapid remapping following spatial cue changes is not always observed but rather depends on several factors, including prior learning experience (Bostock et al. Moreover, if there are sudden shifts from one spatial context to another, the hippocampus spontaneously "flickers" back to the original context representation (Jezek, Henriksen, Treves, Moser, & Moser, 2011). Remapping thus does not simply reflect changes to the perceived spatial cue constellation but rather reflects contextual memory. Situational cues Task and motivational demands strongly influence the firing of hippocampal neurons (Frank, Brown, & Wilson, 2000; Gothard, Skaggs, & McNaughton, 1996; Hampson, Simeral, & Deadwyler, 1999; Kobayashi, Nishijo, Fukuda, Bures, & Ono, 1997; Lee, LeDuke, Chua, McDonald, & Sutherland, 2018; Markus et al. In an even more striking demonstration of the impact of situational context cues, Kelemen and Fenton (2010) trained rats to avoid two shock zones in a rotating disk- shaped arena; one zone was stationary relative to the larger room frame and the other rotated with the arena. Some neurons had place fields that were stationary relative to the broader room framework, while other fields rotated along with the local cues of the rotating arena (figure 19. Thus, the hippocampus held distinct representations of two situational contexts in the same spatial context, one What Contextual Cues Induce Hippocampal Remapping Spatial cues Remapping is induced by spatial cue changes, such as when the walls of a familiar testing arena are replaced with walls of a different color (Bostock et al. The conditions under which remapping (sometimes called global or complex remapping) versus rate remapping are observed are not currently well understood, but whereas global remapping may relate more to contextual changes, rate remapping may reflect noncontextual, nonspatial influences on hippocampal representations (Leutgeb et al. Temporal cues Circadian rhythms modulate the firing rates of hippocampal neurons (Munn & Bilkey, 2012), but whether changes in behaviorally relevant biological times of day induce remapping is less well studied. Greater evidence supports the idea that the hippocampus encodes the relative temporal context in which stimuli are learned and remaps between event sequences with dif ferent temporal structures. Temporal sequence information is represented by hippocampal cells that encode successive moments during a temporal gap between events (MacDonald, Lepage, Eden, & Eichenbaum, 2011; Sakon, Naya, Wirth, & Suzuki, 2014), even for sequences devoid of specific discrete cues (Farovik, Dupont, & Eichenbaum, 2010; Hales & Brewer, 2010; Meck, Church, & Olton, 1984; Moyer, Deyo, & Disterhoft, 1990; Staresina & Davachi, 2009). Critically, many hippocampal neurons sensitive to temporal information remap (or "retime") when the main temporal parameter of a task is altered (figure 19. Many neurons were active at both time points but not reactivated in a different context, indicating that hippocampal context representations remain stable over weeks. Inactivation of the hippocampus prior to context preexposure also eliminates the effect of preexposure in contextual fear- conditioning paradigms (Matus-Amat, Higgins, Barrientos, & Rudy, 2004), suggesting that preexposure allows the hippocampus to form a contextual representation reflecting stable cues. Likewise, spatial cues that are previously experienced as unstable have little control over place fields (Knierim, Kudrimoti, & McNaughton, 1995). Despite the stability of hippocampal context representations, hippocampal population activity changes over time in the presence of the same spatial and situational cues (Mankin et al. Ziv and colleagues (2013) used calcium imaging to monitor the activity of hundreds of hippocampal neurons in mice over a 45- day period. Indeed, the overlap between hippocampal populations activated by two distinct spatial contexts acquired within a day is higher than when separated by a week (Cai et al. Hippocampal contextual representations do not reflect discrete cues Hippocampal lesions selectively impair contextdependent learning in rodents, but not conditioned responses to discrete cues such as a tone, during both episodic (Kim & Fanselow, 1992; Phillips & LeDoux, 1992; Selden, Everitt, Jarrard, & Robbins, 1991) and spatial (Pearce, Roberts, & Good, 1998) memory tasks. Human patients with hippocampal damage likewise have greater deficits in memory for contextual associations compared to recall or recognition of discrete cues and events (Giovanello, Verfaellie, & Keane, 2003; Holdstock, Mayes, Gong, Roberts, & Kapur, 2005; Mayes, Holdstock, Isaac, Hunkin, & Roberts, 2002; Turriziani, Fadda, Caltagirone, & Carlesimo, 2004). For example, Tayler and colleagues (2013) used genetically engineered mice that express a long-lasting marker of neural activity to compare the hippocampal population active at the time of initial exposure to a context with Julian and Doeller: Context in Spatial and Episodic Memory 223 Importantly, consistent with these lesion and neuroimaging results, changes to discrete spatial cues do not always elicit remapping (Cressant, Muller, & Poucet, 1997; Deshmukh & Knierim, 2013; figure 19. Hippocampal representations reflect reliable organization of contextual cues When spatial and episodic cues are hierarchically structured, hippocampal neurons differentiate between such cues using a hierarchical coding scheme (Takahashi, 2013). McKenzie and colleagues (2014) recorded hippocampal neurons while rats explored two rooms containing two objects (A and B) located in either of two positions (figure 19. The rats subsequently learned new room- object-reward contingencies using a second object set (C and D) within the same rooms. At the next level, the population responded similarly to objects at similar positions independent of the valence, and so forth. Thus, the hippocampus can represent cues using a hierarchical coding scheme in which each kind of response represents a subset of the responses at the next highest level of coding. Broadly, this suggests that the hippocampus represents contextual cues in a manner that reflects the reliable organization of those cues. Interestingly, rather than a distinct hippocampal ensemble representing each different context, this would imply that hippocampal neurons do not remap randomly across contexts; rather, the similarity between dif ferent hippocampal context representations may reflect the similarity in across- context relational cue structure, thus enabling across- context behavioral predictions. Consistent with this idea, when only a subset of cues change across contexts, partial remapping can occur in which the place fields of only a proportion of neurons remap (Anderson & Jeffery, 2003). Thus, the mice learned to fear an artificially reactivated representation of the original context even though they had never been shocked there. Since hippocampal activity elicited by stimulation acted as a ser viceable substitute for contextual cues-akin to how recalling the original learning context at retrieval eliminates contextual interference effects-hippocampal context representations mediate context- dependent behav ior. Despite this growing evidence that hippocampal activity is sufficient to induce context- dependent behavior, there is conflicting evidence regarding whether remapping is necessary for contextual memory under more naturalistic conditions. On the one hand, Kennedy and Shapiro (2009) observed remapping due to changes in motivational state (hunger vs. On the other hand, a consistent relationship between remapping and context- dependent behav ior is not always found. Jeffery and colleagues (2003) trained rats to locate a reward in a chamber with black walls. When the wall color was changed to white, the rats still accurately chose the rewarded location despite the fact that the change in wall color induced remapping. This disconnect could have been due to the fact that behavior in this case was guided by discrete cues. Understanding the link between remapping and contextual memory is a critical area for future research. Context Recognition Inputs to the Hippocampus For context to influence memory, an agent must first recognize the cues that denote the current context. This context recognition process is cognitively dissociable from other aspects of spatial memory (Julian, Keinath, Muzzio, & Epstein, 2015). Since the hippocampus mediates both the contextual memory, as well as the recall, of locations, events, or items within a single context (Eichenbaum, Yonelinas, & Ranganath, 2007; Keinath, Julian, Epstein, & Muzzio, 2017; Redish & Touretzky, 1998; Ranganath, 2010), this raises the possibility that context recognition is performed upstream of the hippocampus itself. On the one hand, lesions of the entire entorhinal region produce contextual memory deficits that Hippocampal Context Representations and Behavior If the hippocampus mediates contextual memory, we would expect a link between hippocampal population activity and context- dependent behav ior. Striking demonstrations of this link come from studies using optogenetics to stimulate hippocampal populations (Liu et al.

This would be a possibility if list 1 items had also intruded into list 2 hypertension uncontrolled icd 9 code generic 0.25 mg digoxin with visa, but this was not the case blood pressure in children order digoxin with paypal, supporting the notion that the destabilized memory was selectively susceptible to memory updating via reconsolidation blood pressure patch purchase 0.25 mg digoxin amex. The above study assumed that the reminder cue successfully elicited memory reactivation but did not explicitly test for it arteria meningea purchase genuine digoxin on-line. In a more recent study (Chan & LaPaglia arteria axillaris purchase 0.25 mg digoxin mastercard, 2013) arteriografia purchase digoxin without prescription, researchers verified memory reactivation by eliciting recall. The study employed two experiments whereby participants viewed a movie about a fictional terrorist attack followed by memory reactivation via a recall test either 20 minutes or 48 hours later. Control participants performed a distractor task (a computer game) in lieu of memory reactivation. Postreactivation, or control, participants listened to an audio recount of the terrorist attack, but the recording misrepresented several details. During a memory test either 20 minutes or 24 hours later, participants showed impaired memory for details that were misrepresented, but only if reactivation of the movie preceded the audio recording. This it yet another example of memory updating that demonstrates the malleability of a reactivated memory in the face of new information. Within their second experiment, the authors assessed the degree of specificity needed for new information to update a reactivated memory. To this end, they presented the postreactivation misinformation as part of a story line unrelated to the original movie. This manipulation did not affect the memory of the initial account, suggesting that declarative memory may only become Orederu and Schiller: the Dynamic Memory Engram Life cycle 279 eligible for update if the new information is highly specific to the original memory. The authors further explain that this requirement for specificity is the reason our declarative memories are not constantly modified by new pieces of information encountered during daily life. Emotional memory While many manipulations that target emotional memory reconsolidation in animal models are not suitable for use in humans, there is a pharmacological agent that modifies memories in animals and is safe for human use: propranolol. Propranolol acts through beta-receptor antagonism to regulate the noradrenergic system, which is involved in the consolidation and reconsolidation of emotional memories. In rodents, propranolol has varying influence on memory modification, depending on the memory subtype. In appetitive- conditioning tasks, propranolol with reactivation decreased the self-administration of cocaine and sucrose, with modest effects on reducing alcohol administration. In humans, propranolol with memory reactivation decreased emotional responses to threat memories in healthy controls as well as anxiety patient populations. Similarly, in tasks of appetitive drug- cue associations, recall for emotional memory components was impaired in participants who received propranolol with reactivation, indicating that beta-receptor antagonism may specifically reduce the emotional affect associated with a memory. These results and others illustrate the therapeutic promise for using propranolol to modify maladaptive memories, although the specific clinical applications might be more complex. Some studies have found no effect of propranolol in patient populations, while others have demonstrated efficacy with multiple doses and prereactivation administration. Alongside discoveries using pharmacological agents, scientists have also found noninvasive means to update emotional memories. Conditioned threat memory can be diminished with a behavioral extinction paradigm applied during the reconsolidation window in both rats (Monfils, Cowansage, Klann, & LeDoux, 2009) and humans, with humans showing attenuated threat responding even one year later (Schiller et al. Extinction during reconsolidation may be regarded as a form of updating the initial memory with the "safe" association conveyed during extinction. Similar threat response attenuation was demonstrated using counterconditioning (replacing a negative cue association with a positive one) during reconsolidation and when participants played a computer game following the reminder, which is thought to funnel cognitive resources away from restabilization-thereby disrupting it. These findings support a model of therapeutic reconsolidation with the potential to offer lasting treatment options to patients with anxiety-based psychiatric conditions rooted in maladaptive emotional memories. As appetitive associations are also susceptible to noninvasive interventions during reconsolidation, psychiatric disorders rooted in dysfunctional reward circuitry, such as addiction, are also likely to benefit from reconsolidation-based therapeutics (for a review, see Lee, Nader, & Schiller, 2017). Potentiating reconsolidation Future therapies that target reconsolidation must be careful to modulate memories in the appropriate direction, as experimental manipulations to impair reconsolidation coexist with manipulations that can enhance it. Memory enhancement, though, has therapeutic potential in its own right, as it would be desirable to enhance adaptive memories. Stress has repeatedly been found to enhance hippocampus- dependent memory in animal models (Maroun & Akirav, 2008), as well as in humans. Another study (Coccoz, Sandoval, Stehberg, & Delorenzi, 2013) tested whether declarative memory could be enhanced during the reconsolidation of a forgotten memory. Six days after training in a cue- syllable associative task, a control group of participants showed poor recall for the memory. The authors note that their employed stressor was milder than that of other studies and that a more intense stressor may have enhanced memory even at day 20. The specific type of stress may also have an impact on the direction of reconsolidation effects, as both the elevated platform task (the rat is placed on an elevated platform in a brightly lit room) and context unfamiliarity (the rat is not exposed to the training context prior to training) induce increased glucocorticoid secretion, but the two tasks enhance and impair the reconsolidation of object recognition memory, respectively (Maroun & Akirav, 2008). Stress, though, is not the only mechanism that can enhance the reconsolidation of declarative memory. For example, low, but not high, doses of nicotine administered during the reconsolidation of object recognition enhanced memory in rats (Tian, Pan, & You, 2015), covert variations in sensorimotor demands enhanced motor memory in humans (Wymbs, Bastian, & Celnik, 2016), and transcranial direct current stimulation (a noninvasive method of electrically stimulating the brain using electrodes placed on the scalp) enhanced declarative memory when applied during consolidation and reconsolidation in humans (Javadi & Cheng, 2013). Life of the Engram Postreconsolidation In a typical reconsolidation study, a memory is acquired on day 1 then reactivated and manipulated on day 2. To assess whether day 2 had a lasting effect on the target memory, researchers determine the strength and accessibility of the memory trace by presenting a reminder cue. Probing for recall is a logical method for memory testing, but it is important to keep in mind that the seemingly simple act of stimulating memory retrieval requires reactivation, which makes the memory again susceptible to a number of fates, including destabilization. Each reactivation, even those that occur during memory testing, can initiate a cascade of events. In the days, weeks, and months following memory acquisition, consolidation, reactivation, destabilization, and restabilization, even more still happens to the engram. Thus far we have discussed synaptic, or cellular, consolidation and reconsolidation, which refer to changes at the level of the synapse occurring minutes to hours after learning. Systems consolidation is a process driven by synaptic consolidation but specifically refers to circuit-level changes that convert a memory from an initial hippocampus- dependent state to a hippocampus-independent state. Systems consolidation was discovered when researchers found that lesioning the hippocampus 24 hours postlearning disrupted a contextual threat memory, showing that intact hippocampus function is necessary for memory retrieval. Lesioning the hippocampus 28 days after memory acquisition, however, did not affect memory recall (Kim & Fanselow, 1992). Thus, the hippocampus was determined to be involved in initial synaptic consolidation but with time, the memory is distributed to a range of cortical memory storage sites. In the tradition of reconsolidation mechanisms mirroring those of consolidation, scientists have additionally uncovered evidence for systems reconsolidation. Though many studies report the hippocampal independence of older memories, the authors noted, those studies conflate the memory state and age and fail to account for the fact that older memories are more likely to be in an inactive state. The researchers dissociated hippocampal involvement in active memories from incidental associations with memory age by reactivating remote memories in rats prior to lesioning their hippocampi. They found that hippocampal lesions caused amnesia only if the memory was reactivated prior to the lesion, indicating that reactivation caused the memory to become hippocampus- dependent (Land, Bunsey, & Riccio, 2000). Debiec and colleagues (2002) later used a contextual threat- conditioning paradigm to directly probe systems reconsolidation using a task known to rely on the hippocampus for initial memory encoding and consolidation. Their results again revealed that a hippocampus-independent consolidated contextual threat memory could be made hippocampus- dependent by reactivating the memory, supporting the notion that hippocampal dependence is a function of memory state. Memory researchers have uncovered several pharmacological and behavioral manipulations that relieve the symptoms of psychopathologies rooted in maladaptive memory processing. Patient studies in reconsolidation aim to repurpose these manipulations to go deeper than symptom relief and modify the maladaptive Orederu and Schiller: the Dynamic Memory Engram Life cycle 281 memory itself. A handful of studies have directly assessed the ability to harness reconsolidation to modify pathological memory associations (for reviews, see Exton-McGuinness & Milton, 2018; Kroes, Schiller, LeDoux, & Phelps, 2016; Lee, Nader, & Schiller, 2017). After this reactivating and destabilizing procedure, participants viewed alcohol cues paired with disgusting images in a counterconditioning protocol that lead to a later reduction in cueinduced craving. Cravings also diminished in two other studies examining participants with heroin use disorder and participants who smoke cigarettes. A retrievalextinction procedure led to reduced craving 24 hours later and at a six-month follow-up among patients with heroin use disorder and a one-month follow-up among patients with tobacco use disorder. Participants with a spider phobia also experienced lasting clinical improvements in response to a retrieval- extinction protocol and a reactivation-propranolol protocol, as evidenced by increased approach behav ior toward spiders 24 hours after the extinction session as well as six months and one year later, respectively. Two other studies using retrieval- extinction protocols to modify behavioral expression in spider phobics did not show conclusive evidence of memory modification resulting from reconsolidation manipulation. Together, these experiments demonstrate the potential for therapeutic reconsolidation but also indicate the necessity for clarification of the parameters that reliably correspond to significant clinical improvements. Challenges to Reconsolidation Theory the validity of any scientific theory must be challenged by considering alternative explanations for experimental observations. Accordingly, some scientists argue that the changes in behavioral expression thought to reflect memory modification during reconsolidation could be attributed to other processes that do not modify the memory. The question of whether retrograde amnesia constitutes a storage failure or retrieval failure is at the heart of this reconsolidation debate. If perceived memory modification results from a storage failure, amnesia occurs because a destabilized memory cannot be successfully restabilized. However, if the memory engram remains intact and does not undergo modification, amnesia must occur because the participant no longer has access to the engram, constituting a retrieval failure. In the case of a retrieval failure, the manipulation does not modify the memory itself but modifies the ability for a retrieval cue to successfully access the memory. The support for retrieval failure stems largely from studies that have reversed retrograde amnesia. This memory restoration occurred despite the reversal of synaptic plasticity in engram cells (increased potentiation and dendritic spine density) following blocked reconsolidation (Ryan, Roy, Pignatelli, Arons, & Tonegawa, 2015). This suggests that retrograde amnesia may be the result of state-dependent learning rather than a failure of memory restorage. State- dependent learning and reconsolidation theory, however, are not necessarily mutually exclusive. Additionally, retrograde amnesia may be a shared end point for several neural processes, including disrupted reconsolidation and state- dependent learning. Summary Though memory was once thought to be immutable following consolidation, neuroscientists have found that memory fluctuates between active and inactive states that differentially permit modification. Only an active memory, whether newly acquired or reactivated, can undergo memory destabilization, a neural process that returns the memory to its unstable state through a cascade of molecular, cellular, and genetic events. Once destabilized, the memory can be diminished if restabilization is interrupted, enhanced by potentiating manipulations, or updated with new information. Though a reactivated memory is subject to multiple fates, the last few decades have been marked by increased interest in the specific sequence of memory reactivation, destabilization, and restabilization-namely, due to the immense 282 Memory potential it poses for the treatment of psychopathologies marked by maladaptive memory processing. However, the invasive nature of many experimental manipulations used in studying memory modification prohibits their use in humans, complicating the translation of animal findings to humans. Researchers are currently developing strategies to circumvent this obstacle and have already made strides in uncovering knowledge on memory modifications in humans. Mechanisms of memory stabilization: Are consolidation and reconsolidation similar or distinct processes The enhancement of reconsolidation with a naturalistic mild stressor improves the expression of a declarative memory in humans. The temporal dynamics of enhancing a human declarative memory during reconsolidation. Human memory reconsolidation: A guiding framework and critical review of the evidence. Reconsolidation blockade for the treatment of addiction: Challenges, new targets, and opportunities. Psychedelics and reconsolidation of traumatic and appetitive maladaptive memories: Focus on cannabinoids and ketamine. Mechanisms governing the reactivation- dependent destabilization of memories and their role in extinction. Integration of new information with active memory accounts for retrograde amnesia: A challenge to the consolidation/reconsolidation hypothesis Reconsolidation of episodic memories: A subtle reminder triggers integration of new information. Reactivation of recall-induced neurons contributes to remote fear memory attenuation. Arousal and stress effects on consolidation and reconsolidation of recognition memory. Retrograde amnesia produced by electroconvulsive shock after reactivation of a consolidated memory trace. Extinction-reconsolidation boundaries: Key to persistent attenuation of fear memories. Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. On the role of hippocampal protein synthesis in the consolidation and reconsolidation of object recognition memory. Previous editions focused on attention in isolation, but the focus of research has shifted over recent years. The cognitive neuroscience of working memory has become a large and relatively mature field, and working memory is strongly intertwined with attention, so it made sense to combine attention and working memory in the same section. Interestingly, although we invited the chapter authors to contribute a chapter on attention or working memory, most of the authors wrote chapters on attention and working memory. A second exciting innovation for our section is that we include, for the first time, a chapter on the development of attention and working-memory functions (by Scerif).

These two results support the provocative suggestion that the focus of attention may not simply be a monolithic process applied to attended items blood pressure drop symptoms generic digoxin 0.25 mg mastercard. It may instead comprise at least two complementary but distinct facets of neural activity blood pressure line chart buy discount digoxin 0.25mg on line. In the attention task pulse pressure range elderly discount digoxin online visa, subjects instead attended to the positions of the colors in anticipation of an occasional brief target whose orientation had to be discriminated prehypertension stage 1 buy cheap digoxin 0.25mg on-line. In line with the expectation that both tasks would recruit spatial attention to the relevant side arteria3d viking pack cheap 0.25 mg digoxin, both tasks produced highly reliable modulations of sustained contralateral alpha power blood pressure medication long term effects 0.25mg digoxin overnight delivery. These results provide initial evidence that these two neural measures of the focus of attention may play distinct roles: one that represents objects in active memory and another that provides a map of currently prioritized space (see also Bae & Luck, 2018). Alpha and Prioritized Space the modulations of contralateral alpha power in the Hakim et al. Moreover, recent work has demonstrated that alpha topography precisely tracks the relevant position in a hemifield, not just the attended side of space. Thus, the spatial information encoded in alpha activity has the graded character that is a hallmark of sensory representations of space. These findings suggest that at least two distinct neural signals track items within the focus of attention. Thus, the neural activity supporting the focus of attention reflects a collaboration between multiple processes that play distinct roles in online memory. When attention returned to that item, the neural activity tracking that item returned. Here, information is stored in a passive manner that enables the rapid reactivation of recently attended information. This mode of storage is less metabolically demanding and may be particularly well suited for guiding comparisons between new inputs and recently attended ones. Indeed, more recent studies have shown that transcranial magnetic stimulation (Rose et al. On the one hand, the recent work on activity- silent memory has provided an exciting new window into the neural mechanisms that can support the retention of information over brief delays. On the other hand, there is room for debate regarding the most productive way to position these activity- silent phenomena within a taxonomy of memory. Is a rapid shift of synaptic weights-in the absence of active neural signals-best understood as working memory One might presume so, given that behavioral tests show that subjects can still access the target information following the short delay. Thus, activity- silent memories have typically been referred to as unattended memory items. A central virtue of the embedded process models is their acknowledgment that both active and passive aspects of memory are critical for virtually any complex cognitive task. While it has been postulated that the sustained maintenance of activity- silent representations may be contingent on current behavioral relevance. Likewise, the contents of past trials shape the responses to items in the present, even though past trials are completely irrelevant. Thus, given that recently attended items often exert influence when they are behaviorally irrelevant (Awh, Belopolsky, & Theeuwes, 2012), more work is needed to determine the relationship between activity- silent representations and voluntary control. Our intent is not to promote endless debate over how to label various memory phenomena. Even amid any ongoing controversy regarding the best way to categorize dif ferent memory phenomena, there is nevertheless a consensus that we should push forward with the effort to link robust behavioral indices of memory function with clear models of the underlying neural processes. Delineating resetting and updating in visual working memory based on the object-to-representation correspondence. Neural and behavioral evidence for an online resetting process in visual working memory. Models of Working Memory: Mechanisms of Active Maintenance and Executive Control, 20, 506. Neural mea sures of individual differences in selecting and tracking multiple moving objects. Long-term working memory as an alternative to capacity models of working memory in everyday skilled per for mance. Retrospective cues mitigate information loss in human cortex during working memory storage. Alphaband activity reveals spontaneous representations of spatial position in visual working memory. The topography of alpha- band activity tracks the content of spatial working memory. The role of alpha-band brain oscillations as a sensory suppression mechanism during selective attention. Quantity, not quality: the relationship between fluid intelligence and working memory capacity. Individual differences in visual working memory capacity: Contributions of attentional control to storage. Coarse-to-fine construction for high-resolution representation in visual working memory. Dissecting the neural focus of attention reveals distinct processes for spatial attention and object-based storage in visual working memory. Contralateral delay activity provides a neural mea sure of the number of representations in visual working memory. Neural evidence for a distinction between short-term memory and the focus of attention. Shape and color conjunction stimuli are represented as bound objects in visual working memory. Electrophysiological mea sures of maintaining representations in visual working memory. Are representations in working memory distinct from representations in long-term memory Mental rotation requires visual short-term memory: Evidence from human electric cortical activity. Awh and Vogel: Online and Off-Line Memory States in the Human Brain 355 Stokes, M. The nature of individual differences in working memory capacity: Active maintenance in primary memory and controlled search from secondary memory. Working memory and fluid intelligence: Capacity, attention control, and secondary memory retrieval. Modulations in oscillatory activity with amplitude asymmetry can produce cognitively relevant event-related responses. Neural activity predicts individual differences in visual working memory capacity. Neural mea sures reveal individual differences in controlling access to working memory. Visual- spatial attention aids the maintenance of object representations in visual working memory. Synchronizing the activity of neural populations can facilitate communication within the ensemble. When synchronized in phase with one another, neurons are excitable (or not) at the same time. When they are both in an excitable state, spikes from one neuron will have a greater impact on the other, facilitating communication. On the other hand, if neurons are out of sync or anticorrelated, one set of neurons may be spiking when another set is in a low state of excitement, hindering the impact of spikes and thus limiting communication between them. The advantage of forming ensembles via rhythmic synchrony is that they are flexible. Ensembles can be formed, discarded, and then reformed, all by changing the pattern of synchrony without needing to change the physical structure. A, Working memory representations are distributed across the brain, including in sensory regions, parietal regions, and prefrontal regions, as well as subcortical regions, such as the basal ganglia and the thalamus. Synchrony within and between different brain regions is thought to help organize the distributed representation into a cohesive representation. B, Working memory is represented in the sustained neural activity of prefrontal cortex neurons. For example, a prefrontal cortex neuron persistently responds when a monkey remembers a stimulus presented to the left of fixation (third column) compared to when the same stimulus was presented to the right, up, or down (other columns, from left to right). Cross-temporal correlation shows that, across a population of prefrontal cortex neurons, neural activity at one time point (x-axis) is not well correlated with activity at other time points (y-axis). In particular, correlation is low between the response to the stimulus presentation (shaded gray on x-axis) and memory delay. Despite the dynamics seen in C, a mnemonic subspace exists in which different memories (indicated by different colors) can be stably decoded. This is thought to be due to recurrent connections between the neurons that belong to the same ensemble. The idea is that once activity passes a threshold, there is enough recurrence to sustain its activity. In this model, neurons are topographically arranged around a ring according to their selectivity-nearby neurons share similar selectivity. Local recurrent connections then sustain initial inputs into the ring, leading to a "bump" of activity, while more distal inhibitory connections stabilize the memory in place. This leads to a persistent attractor state, corresponding to a specific pattern of activity. However, recent work is beginning to challenge that (reviewed in Lundqvist, Herman, & Miller, 2018; Stokes, 2015). Much of the prior evidence for persistent spiking comes from studies that averaged spiking across time and trials. While this shows that the average spike rate of neurons increases over the delay, it masks the details of the spiking itself. Watanabe and Funahashi (2014) trained monkeys on an oculomotor delayed- saccade task that required memory for the location of a saccade target. When the delay duration is fixed, robust spiking may only emerge late in the delay. The resulting "ramp" in neural activity may reflect preparation for the upcoming memory probe, suggesting the activity spiking is a readout, not memory, mechanism. This can be evaluated by testing if a decoder trained on activity at one time in the trial can decode memories at other times. It is possible, however, to find a linear combination of neurons that will maintain a stable code, "a stable subspace" (figure 31. However, it is important to note that this has been demonstrated with "empty" delays without additional inputs or distractions. Decoders trained before additional inputs do not perform well following 358 Attention and Working Memory them. This change in code is consistent with mixed selectivity-individual neurons sensitive to the combination of multiple behavioral conditions and items (Rigotti et al. In other words, the spikes leave an "impression" in networks that preserve the memory of the activity. Indeed, spiking activity can produce fast synaptic enhancement that last hundreds of milliseconds (Wang et al. Memories can be maintained over a longer time scale by "refreshing" the synaptic weight changes with occasional spiking. Such activity- silent representations have functional advantages over persistent spiking. Memories held by persistent spiking alone can be labile because they are lost when activity is disrupted. If there is any overlap in the ensembles/attractor states, they tend to meld into one. Plus, neurons optimize information when they spike sparsely and in bursts, not persistently. Activity- silent models predict content- dependent changes in network connections. It is difficult to directly test this prediction, as it is difficult to record from a pair of monosynpatically connected neurons. A further prediction is that neural responses to a new input should depend on the information already encoded in synaptic weights. If too many items are simultaneously held, the requirement to refresh the synapses causes a buildup of interference due to competition for the limited time available for the refresh. Unlike long-term memory, which has enough capacity to hold a lifetime of experiences and knowledge, we can only hold very few items "in mind" simultaneously. Individual capacity varies from one to seven and is highly correlated with fluid intelligence, reflecting that capacity limits are a fundamental restriction in cognition (Fukuda, Vogel, Mayr, & Awh, 2010). This makes sense: the more thoughts that can be simultaneously held and manipulated, the more associations, connections, and relationships can be made and therefore the more sophisticated a thought can be. Or do we try to take in as much information as possible, eventually spreading ourselves too thin In these tasks, subjects must remember a screen with a variable number of objects (such as colored squares). Then, after a delay of a few seconds, the subjects see a second "test" screen of objects. In contrast, the resource model predicts that increasing memory load should reduce the information about any single item.

Buy digoxin 0.25mg on-line. Heat can be harmful to people on blood pressure medication.