SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4247296973> ?p ?o ?g. }

Showing items 1 to 47 of 47 with 100 items per page.

W4247296973 abstract "Full text Figures and data Side by side Abstract eLife digest Introduction Results Discussion Materials and methods References Decision letter Author response Article and author information Metrics Abstract To successfully guide limb movements, the brain takes in sensory information about the limb, internally tracks the state of the limb, and produces appropriate motor commands. It is widely believed that this process uses an internal model, which describes our prior beliefs about how the limb responds to motor commands. Here, we leveraged a brain-machine interface (BMI) paradigm in rhesus monkeys and novel statistical analyses of neural population activity to gain insight into moment-by-moment internal model computations. We discovered that a mismatch between subjects’ internal models and the actual BMI explains roughly 65% of movement errors, as well as long-standing deficiencies in BMI speed control. We then used the internal models to characterize how the neural population activity changes during BMI learning. More broadly, this work provides an approach for interpreting neural population activity in the context of how prior beliefs guide the transformation of sensory input to motor output. https://doi.org/10.7554/eLife.10015.001 eLife digest The human brain is widely hypothesized to construct “inner beliefs” about how the world works. It is thought that we need this conception to coordinate our movements and anticipate rapid events that go on around us. A driver, for example, needs to predict how the car should behave in response to every turn of the steering wheel and every tap on the brake. But on icy roads, these predictions will often not reflect how the car would behave. Applying the brakes sharply in these conditions could send the car skidding uncontrollably rather than stopping. In general, a mismatch between one’s inner beliefs and reality is thought to cause errors and accidents. Yet this compelling hypothesis has not yet been fully investigated. Golub et al. investigated this hypothesis by conducting a “brain-machine interface” experiment. In this experiment, neural signals from the brains of two rhesus macaques were recorded using arrays of electrodes and translated into movements of a cursor on a computer screen. The monkeys were then trained to mentally move the cursor to hit targets on the screen. The monkeys’ cursor movements were remarkably precise. In fact, the experiment showed that the monkeys could internally predict their cursor movements just as a driver predicts how a car will move when turning the steering wheel. These findings indicate that the monkeys have likely developed inner beliefs to predict how their neural signals drive the cursor, and that these beliefs helped coordinate their performance. In addition, when the monkeys did make mistakes, their neural signals were not entirely wrong—in fact they were typically consistent with the monkeys’ inner beliefs about how the cursor moves. A mismatch between these inner beliefs and reality explained most of the monkeys’ mistakes. The brain constructs such inner beliefs over time through experience and learning. To study this learning process, Golub et al. next conducted an experiment in which the cursor moved in a way that was substantially different from the monkey’s inner beliefs. This experiment uncovered that, during the course of learning, the monkey’s inner beliefs realigned to better match the movements of the new cursor. Taken together, this work provides a framework for understanding how the brain transforms sensory information into instructions for movement. The findings could also help improve the performance of brain-machine interfaces and suggest how we can learn new skills more rapidly and proficiently in everyday life. https://doi.org/10.7554/eLife.10015.002 Introduction Even simple movements, like reaching to grasp a glass of water, require dozens of muscles to be activated with precise coordination. This precision is especially impressive in light of sensory feedback delays inherent to neural transmission and processing: when we make a swift arm movement, the brain only knows where the arm was a split second ago, not where it currently is. To generate the desired movement, it is widely believed that we form internal models that enable selection of appropriate motor commands and prediction of the outcomes of motor commands before sensory feedback becomes available (Crapse and Sommer, 2008; Shadmehr et al., 2010). Mechanistic studies have made important progress toward identifying the neural circuits that implement internal models in sensory (Komatsu, 2006; Kennedy et al., 2014; Schneider et al., 2014), vestibular (Laurens et al., 2013), and motor (Sommer, 2002; Ghasia et al., 2008; Keller and Hahnloser, 2009; Azim et al., 2014) systems. In parallel, psychophysical studies have demonstrated the behavioral correlates of these internal models (Shadmehr and Mussa-Ivaldi, 1994; Wolpert et al., 1995; Thoroughman and Shadmehr, 2000; Kluzik et al., 2008; Mischiati et al., 2015) and the behavioral deficits that result from lesions to corresponding brain areas (Shadmehr and Krakauer, 2008; Bhanpuri et al., 2013). Together with studies showing neural correlates of internal models (Sommer, 2002; Gribble and Scott, 2002; Ghasia et al., 2008; Mulliken et al., 2008; Keller and Hahnloser, 2009; Green and Angelaki, 2010; Berkes et al., 2011; Laurens et al., 2013), these previous studies have provided strong evidence for the brain’s use of internal models. These internal models are presumably rich entities that reflect the multi-dimensional neural processes observed in many brain areas (Cunningham and Yu, 2014) and can drive moment-by-moment decisions and motor output. However, to date, most studies have viewed internal models through the lens of individual neurons or low-dimensional behavioral measurements, which provides a limited view of these multi-dimensional neural processes (although see Berkes et al., 2011). Here, we address these limitations by extracting a rich internal model from the activity of tens of neurons recorded simultaneously. The key question that we ask is whether such an internal model can explain behavioral errors that cannot be explained by analyzing low-dimensional behavioral measurements in isolation. We define an internal model to be one’s inner conception of a motor effector, which includes one’s prior beliefs about the physics of the effector as well as how neural commands drive movements of the effector. When we extract a subject’s internal model, we seek a statistical model of the effector dynamics that is most consistent with the subject’s neural commands. Interpreting high-dimensional neural activity through the lens of such an internal model offers insight into how one’s prior beliefs about the effector affect the transformation of sensory inputs into population-level motor commands on a timescale of tens of milliseconds. To date, it has been difficult to identify such an internal model due the complexities of non-linear effector dynamics and multiple sensory feedback modalities, the need to monitor many neurons simultaneously, and the lack of an appropriate statistical algorithm. To overcome these difficulties, we leveraged a closed-loop brain-machine interface (BMI) paradigm (Figure 1A) in rhesus monkeys, which translates neural activity from the primary motor cortex (M1) into movements of a computer cursor (Green and Kalaska, 2011). A BMI represents a simplified and well-defined feedback control system, which facilitates the study of internal models (Golub et al., 2016). In particular, the BMI mapping from neural activity to movements is completely specified by the experimenter and can be chosen to define linear cursor dynamics, the relevant sensory feedback can be limited to one modality (in this case, vision), and all neural activity that directly drives the cursor is recorded. Figure 1 with 1 supplement see all Download asset Open asset Closed-loop control of a brain-machine interface (BMI) cursor. (A) Schematic view of the brain-machine interface. Subjects produce neural commands to drive a cursor to hit visual targets under visual feedback. (B) Cursor trajectories from the first 10 successful trials to each of 16 instructed targets (filled circles) in representative data sets. Target acquisition was initiated when the cursor visibly overlapped the target, or equivalently when the cursor center entered the cursor-target acceptance zone (dashed circles). Trajectories shown begin at the workspace center and proceed until target acquisition. Data are not shown during target holds. https://doi.org/10.7554/eLife.10015.003 During proficient BMI control, as with other behavioral tasks, subjects make movement errors from time to time. One possible explanation for these errors is that they arise due to sensory or motor “noise” that varies randomly from one trial to the next (Harris and Wolpert, 1998; Osborne et al., 2005; Faisal et al., 2008). Another possibility, which is the central hypothesis in this study, is that a substantial component of movement errors is structured and can be explained by a mismatch between the subject’s internal model of the BMI and the actual BMI mapping. Testing this hypothesis required the development of a novel statistical method for estimating the subject’s internal model from the recorded M1 activity, BMI cursor movements, and behavioral task goals. The internal model represents the subject’s prior beliefs about the physics of the BMI cursor, as well as how the subject’s neural activity drives the cursor. To justify the study of internal models in a BMI context, we first asked whether subjects show evidence of internal prediction during BMI control. Next, we asked whether interpreting M1 activity through extracted internal models could explain movement errors that are present throughout proficient BMI control and long-standing deficiencies in control of BMI movement speed. Finally, because a key feature of internal models is their ability to adapt (Shadmehr et al., 2010), we altered the BMI mapping and asked whether the internal model adapted in a manner consistent with the new BMI mapping. An important distinction that we make relative to previous work is that we are not asking circuit-level questions about how and where in the brain these internal models operate. Rather, we seek a statistical representation of the subject’s prior beliefs about the BMI mapping (i.e., an internal model) that can be used to explain behavioral errors. Although internal models might not reside in M1 (Shadmehr, 1997; Pasalar et al., 2006; Miall et al., 2007; Mulliken et al., 2008; Lisberger, 2009), their computations influence activity in M1. Thus, by examining the moment-by-moment relationship between M1 population activity and task objectives, it may be possible to extract a detailed representation of the subject’s internal model. Results We trained two rhesus monkeys to modulate neural activity to drive movements of a computer cursor to hit targets in a two-dimensional workspace (Figure 1B). The family of BMI mappings that we used is represented by: (1) xt=Axt-1+But+ b where xt is the cursor state (position and velocity), ut comprises the recorded M1 activity, and A, B, and b are the parameters of the BMI mapping. All experiments began with a closed-loop calibration of an intuitive BMI mapping, which was designed to provide proficient control on par with the majority of studies in the field (Serruya et al., 2002; Velliste et al., 2008; Ganguly and Carmena, 2009; Suminski et al., 2010; Hauschild et al., 2012; Ifft et al., 2013; Sadtler et al., 2014). Subjects indeed demonstrated proficient and stable control of the BMI, with success rates of nearly 100%, and movement times on average faster than one second (Figure 1—figure supplement 1). The BMI provides an ideal paradigm for studying internal models because it simplifies several key complexities of native limb control. First, native limb control involves effectors with non-linear dynamics, and the causal relationship between the recorded neural activity and limb movements is not completely understood. In contrast, the causal relationship between recorded neural activity and BMI cursor movements is completely specified by the experimenter (through A, B and b in Equation 1), and can be chosen to be linear (as in Equation 1). Second, native limb control involves multiple modalities of sensory feedback (e.g., proprioception and vision), which makes it difficult for the experimenter to know how the subject combines sources of sensory information. In the BMI, task-relevant sensory feedback is limited to a single modality (vision), which is completely specified by the experimenter (xt in Equation 1). Finally, the neural activity that directly drives the BMI is completely specified by the recorded population activity (ut in Equation 1), whereas typically only a subset of neurons driving limb movements is recorded. We can thereby reinterpret the full set of BMI control signals using an internal model in a more concrete manner than is currently possible with limb movements. Subjects compensate for sensory feedback delays while controlling a BMI Because internal models have not previously been studied in a BMI context, we sought evidence of internal prediction. A hallmark of internal prediction is compensation for sensory feedback delays (Miall et al., 2007; Shadmehr et al., 2010; Farshchiansadegh et al., 2015). To assess the visuomotor latency experienced by a subject in our BMI system, we measured the elapsed time between target onset and the appearance of target-related activity in the recorded neural population (Figure 2A). The delays we measured (100 ms, monkey A; 133 ms, monkey C) are consistent with visuomotor latencies reported in arm reaching studies of single-neurons in primary motor cortex (Schwartz et al., 1988). Next, we asked whether subjects produced motor commands consistent with the current cursor position, which was not known to the subject due to visual feedback delay, or whether motor commands were more consistent with a previous, perceived position (Figure 2B,C and Figure 2—figure supplement 1). If subjects did not compensate for visual feedback delays and aimed from the most recently available visual feedback of cursor position, we would expect errors to be smallest at lags of 100 ms and 133 ms relative to the current cursor position for monkeys A and C, respectively (dashed red lines in Figure 2C). Rather, we found that these error curves had minima at lags close to 0 ms (dashed black lines in Figure 2C), indicating that motor commands through the BMI mapping pointed closer to the targets when originating from the current cursor position than from any previous position. This finding suggests that subjects use an internal model to internally predict the current cursor position. Figure 2 with 1 supplement see all Download asset Open asset Subjects compensate for sensory feedback delays while controlling a BMI. (A) The visuomotor latency experienced by a subject in our BMI system was assessed by measuring the elapsed time between target onset and the first significant (p<0.05) decrease in angular error. If that first decrease was detected τ+1 timesteps following target onset, we concluded that the visuomotor latency was at least τ timesteps (red dashed lines). For both subjects, the first significant difference was highly significant (**p<10-5, two-sided Wilcoxon test with Holm-Bonferroni correction for multiple comparisons; n = 5908 trials; monkey C: n = 4578 trials). (B) Conceptual illustration of a single motor command (black arrows) shifted to originate from positions lagged relative to the current cursor position (open circle). In this example, the command points farther from the target as it is shifted to originate from earlier cursor positions. (C) Motor commands pointed closer to the target when originating from the current cursor position (zero lag) than from outdated (positive lag) cursor positions that could be known from visual feedback alone (**p<10-5, two-sided Wilcoxon test; monkey A: n = 33,660 timesteps across 4489 trials; monkey C: n = 31,214 timesteps across 3639 trials). Red lines indicate subjects’ inherent visual feedback delays from panel A. Shaded regions in panels A and C (barely visible) indicate ± SEM. https://doi.org/10.7554/eLife.10015.005 Because we have not yet explicitly identified the subject’s internal model, motor commands were defined in this analysis using the BMI mapping, which is external to the subject. If the internal model bears similarities to the BMI mapping, it is reasonable to use the BMI mapping as a proxy for the internal model to assess feedback delay compensation. With evidence that subjects engage an internal model during BMI control, we next asked whether we could explicitly identify an internal model from the recorded neural activity. Internal model mismatch explains the majority of subjects’ control errors The BMI mapping, which determines the cursor movements displayed to the subject, provides one relevant, low-dimensional projection of the high-dimensional neural activity. With evidence that subjects use an internal model during closed-loop BMI control, we asked whether mismatch between an internal model and the actual BMI mapping could explain the subject’s moment-by-moment aiming errors. This requires identifying the subject’s internal model, which could reveal a different projection of the high-dimensional neural activity, representing the subject’s internal beliefs about the cursor state. Because of the closed-loop nature of the BMI paradigm, the subject continually updates motor control decisions as new visual feedback of the cursor becomes available. To resolve these effects, the internal model needs to operate on a timescale of tens of milliseconds (in this case, a single timestep of the BMI system) on individual experimental trials. The extraction of such a rich internal model has been difficult prior to this study due to the lack of an appropriate statistical framework. To overcome this limitation, we developed an internal model estimation (IME) framework, which extracts, from recorded population activity, a fully parameterized internal model along with a moment-by-moment account of the internal prediction process (Figure 3A). In the IME framework, the subject internally predicts the cursor state according to: (2) x~t=A~⁢x~t-1+B~ut+b~ Figure 3 with 8 supplements see all Download asset Open asset Mismatch between the internal model and the BMI mapping explains the majority of the subjects’ cursor movement errors. (A) At each timestep, the subject’s internal state predictions (x~t-2, x~t-1, x~t) are formed by integrating the visual feedback (xt-3) with the recently issued neural commands (ut-2, ut-1, ut) using the internal model (A~,B~,b~). We defined cursor states and internal state predictions to include components for position and velocity (i.e., xt=[pt;vt],x~t=[p~t;v~t]). (B) Cursor trajectory (black line) from a BMI trial that was not used in model fitting. Red whisker shows the subject’s internal predictions of cursor state as extracted by IME. The critical comparison is between the actual cursor velocity (vt; black arrow) and the subject’s internal prediction of cursor velocity (v~t; red arrow). (C) Cross-validated angular aiming errors based on IME-extracted internal models are significantly smaller than cursor errors from the BMI mapping (**p<10-5, two-sided Wilcoxon test; monkey A: n = 5908 trials; monkey C: n = 4577 trials). Errors in panel B are from a single timestep within a single trial. Errors in panel C are averaged across timesteps and trials. Errors in panels B and C incorporate temporal smoothing through the definition of the BMI mapping and the internal model, and are thus not directly comparable to the errors shown in Figure 2C, which are based on single-timestep velocity commands needed for additional temporal resolution. Error bars (barely visible) indicate ± SEM. https://doi.org/10.7554/eLife.10015.007 where x~t is the subject’s internal prediction about the cursor state (position and velocity), ut is a vector of recorded neural activity, and A~, B~, and b~ are the parameters of the subject’s internal model. This form of the internal model was chosen to be analogous to the BMI mapping from Equation 1 so that the actual BMI mapping lies within the family of internal models that we consider. Additionally, this formulation aligns with recent studies of skeletomotor (Shadmehr and Krakauer, 2008) and oculomotor (Frens, 2009) control, and a vast literature of control theory (Anderson and Moore, 1990). The primary concept of the IME framework is that, at each timestep, the subject internally predicts the current cursor state by recursively applying Equation 2 (starting from the most recently available sensory feedback) and generates neural activity consistent with aiming straight to the target relative to this internal prediction (see the 'Framework for internal model estimation (IME)' subsection in 'Materials and methods' and Figure 3—figure supplement 1). At each timestep, IME extracts the entire time-evolution of the subject’s internal state prediction using Equation 2 as an internal forward model. This evolution can be visualized in the form of a whisker (Figure 3B) that begins at the cursor position of the most recently available feedback and unfolds according to the extracted internal model. At each new timestep, the subject forms a new internal prediction that incorporates newly received visual feedback. If the internal model exactly matches the BMI mapping, the subject’s internal predictions would exactly match the cursor trajectory. A visualization of an example internal model and BMI mapping is given in Figure 3—figure supplement 2. The central hypothesis in this study is that movement errors arise from a mismatch between the subject’s internal model of the BMI and the actual BMI mapping. The alternative to this hypothesis is that the subject’s internal model is well-matched to the BMI mapping, and movement errors result from other factors, such as “noise” in the sensorimotor system, subjects’ inability to produce certain patterns of neural activity, or subjects disengaging from the task. Our key finding is that recorded neural commands were markedly more consistent with the task goals when interpreted through subjects’ internal models than when viewed through the BMI mapping (Figure 3C). Subjects’ internal models deviated from the actual BMI mappings such that control errors evaluated through extracted internal models were substantially smaller than actual cursor errors: extracted internal models explained roughly 65% of cursor movement errors (70%, monkey A; 59%, monkey C). Although this finding does not preclude other factors (e.g., spiking noise or subject disengagement) from contributing toward movement errors, it does suggest their contribution is substantially smaller than previously thought, due to the large effect of internal model mismatch. We found that the majority of the explanatory power of extracted internal models was in their ability to identify structure in the high-dimensional neural activity (Figure 3—figure supplement 3). This structure was captured in the internal model by the mapping from high-dimensional neural activity to low-dimensional kinematics (B~ in Equation 2), which need not match the BMI mapping (B in Equation 1). Consistent with this finding, internal models fit to low-dimensional behavior rather than high-dimensional neural activity were not able to explain cursor errors (Figure 3—figure supplement 4). That a majority of cursor errors can be explained by mismatch of the internal model is not to say that control through the BMI mapping was poor–in fact control was proficient and stable (Figure 1B and Figure 1—figure supplement 1). Rather, extracted internal models predicted movements that consistently pointed straight to the target, regardless of whether the actual cursor movements did (Figure 4A) or did not (Figure 4B and Figure 4—figure supplement 1) point straight to the target. On most trials, BMI cursor trajectories proceeded roughly straight to the target (Figure 4A). On these trials, internal model predictions aligned with actual cursor movements, resulting in small errors through both the BMI mapping and the extracted internal model. In a smaller subset of trials, actual cursor movements were more circuitous and thus had relatively large errors. Previously, the reason behind these seemingly incorrect movements was unknown, and one possibility was that the subject simply disengaged from the task. When interpreted through the extracted internal model, however, neural activity during these circuitous trials appears correct, suggesting that the subject was engaged but was acting under an internal model that was mismatched to the BMI mapping (Figure 4B and Figure 4—figure supplement 1). In other words, when armed with knowledge of the subject’s internal model, outwardly irrational behavior (i.e., circuitous cursor movements) appears remarkably rational. Across all trials, the majority of neural activity patterns had low or zero error as evaluated through extracted internal models, regardless of whether errors of the actual cursor movements (i.e., through the BMI mapping) were large or small (Figure 4C and Figure 4—figure supplement 2). Figure 4 with 2 supplements see all Download asset Open asset Neural activity appears correct through the internal model, regardless of how the actual cursor moved. (A) Typical trial in which the cursor followed a direct path (black) to the target. Internal model predictions (red whiskers) also point straight to the target. (B) Trial with a circuitous cursor trajectory. Internal model predictions point straight to the target throughout the trial, regardless of the cursor movement direction (same color conventions as in panel A). (C) Timestep-by-timestep distribution of BMI cursor and internal model errors. Neural activity at most timesteps produced near-zero error through the internal model, despite having a range of errors through the BMI mapping. (D) Hypothetical internal model (red) and BMI mapping (black) relating 2D neural activity to a 1D velocity output. This is a simplified visualization of Equations 1 and 2, involving only the B and B~ parameters, respectively. Each contour represents activity patterns producing the same velocity through the internal model (v~, red) or BMI mapping (v, black). Because of internal model mismatch, many patterns result in different outputs through the internal model and the BMI. However, some patterns result in the same output through both the internal model and the BMI (gray line). Here we illustrate using a 2D neural space and 1D velocity space. In experiments with q-dimensional neural activity and 2D velocity, activity patterns producing identical velocities through both the internal model and the cursor span a (q-4)-dimensional space. https://doi.org/10.7554/eLife.10015.016 When cursor trajectories were circuitous, it was not uncommon for some internal model predictions (whiskers) to match the actual cursor movement while others did not, even within the same trial (Figure 4B). Given a single internal model, how can some patterns of neural activity result in whiskers aligned to the cursor trajectory, while others patterns produce whiskers that deviate from the cursor trajectory? This is possible due to mathematical operation of mapping from high-dimensional neural activity patterns to low-dimensional cursor states. Figure 4D provides a conceptual illustration of a simplified BMI mapping: (3) vt=But and a simplified internal model: (4) vt~=B~⁢ut each of which relies only on a mapping (B or B~) from neural activity (ut) to cursor velocity (vt or v~t). We focus on B and B~ here (without considering A, b, A~, and b~ from Equations 1 and 2) because of the aforementioned finding that the majority of the internal model mismatch effect is captured by differences between B and B~ (Figure 3—figure supplement 3). Given a mismatched BMI mapping (black lines) and internal model (red lines), many neural activity patterns will produce different velocities through the BMI mapping versus the internal model. However, a subset of activity patterns (gray line) will produce identical velocities through both the BMI mapping and the internal model. These patterns lie in the nullspace of B-B~ (i.e., solutions to the equation But=B~⁢ut). In the example trials shown in Figure 4A,B and Figure 4—figure supplement 1, internal model predictions (red) that match the actual cursor movement (black) correspond to neural activity patterns along the gray line in Figure 4D. Predictions not matching the cursor movement correspond to neural activity patterns anywhere off the gray line in Figure 4D. Two alternative hypotheses do not explain the effect of internal model mismatch The data presented thus far support our central hypothesis that internal model mismatch is a primary source of movement errors. Next we asked whether it might be possible to have arrived at this result under the alternate hypothesis that the internal model is well-matched to the BMI mapping. We address two specific cases of this alternative hypothesis and show that they do not explain the observed effect of internal model mismatch. First, we explored the possibility that the subject might have a well-matched in" @default.
W4247296973 created "2022-05-12" @default.
W4247296973 date "2015-09-04" @default.
W4247296973 modified "2023-09-24" @default.
W4247296973 title "Decision letter: Internal models for interpreting neural population activity during sensorimotor control" @default.
W4247296973 doi "https://doi.org/10.7554/elife.10015.022" @default.
W4247296973 hasPublicationYear "2015" @default.
W4247296973 type Work @default.
W4247296973 citedByCount "0" @default.
W4247296973 crossrefType "peer-review" @default.
W4247296973 hasBestOaLocation W42472969731 @default.
W4247296973 hasConcept C154945302 @default.
W4247296973 hasConcept C15744967 @default.
W4247296973 hasConcept C169760540 @default.
W4247296973 hasConcept C2775924081 @default.
W4247296973 hasConcept C28427503 @default.
W4247296973 hasConcept C2908647359 @default.
W4247296973 hasConcept C2984127161 @default.
W4247296973 hasConcept C41008148 @default.
W4247296973 hasConcept C71924100 @default.
W4247296973 hasConcept C99454951 @default.
W4247296973 hasConceptScore W4247296973C154945302 @default.
W4247296973 hasConceptScore W4247296973C15744967 @default.
W4247296973 hasConceptScore W4247296973C169760540 @default.
W4247296973 hasConceptScore W4247296973C2775924081 @default.
W4247296973 hasConceptScore W4247296973C28427503 @default.
W4247296973 hasConceptScore W4247296973C2908647359 @default.
W4247296973 hasConceptScore W4247296973C2984127161 @default.
W4247296973 hasConceptScore W4247296973C41008148 @default.
W4247296973 hasConceptScore W4247296973C71924100 @default.
W4247296973 hasConceptScore W4247296973C99454951 @default.
W4247296973 hasLocation W42472969731 @default.
W4247296973 hasOpenAccess W4247296973 @default.
W4247296973 hasPrimaryLocation W42472969731 @default.
W4247296973 hasRelatedWork W1968338851 @default.
W4247296973 hasRelatedWork W2019149755 @default.
W4247296973 hasRelatedWork W2035002319 @default.
W4247296973 hasRelatedWork W2069197020 @default.
W4247296973 hasRelatedWork W2092543968 @default.
W4247296973 hasRelatedWork W2141330654 @default.
W4247296973 hasRelatedWork W2748952813 @default.
W4247296973 hasRelatedWork W2899084033 @default.
W4247296973 hasRelatedWork W2913276869 @default.
W4247296973 hasRelatedWork W2962385765 @default.
W4247296973 isParatext "false" @default.
W4247296973 isRetracted "false" @default.
W4247296973 workType "peer-review" @default.