RESUMEN

Bumblebees (Bombus terrestris) have been shown to engage in string-pulling behavior to access rewards. The objective of this study was to elucidate whether bumblebees display means-end comprehension in a string-pulling task. We presented bumblebees with two options: one where a string was connected to an artificial flower containing a reward and the other presenting an interrupted string. Bumblebees displayed a consistent preference for pulling connected strings over interrupted ones after training with a stepwise pulling technique. When exposed to novel string colors, bees continued to exhibit a bias towards pulling the connected string. This suggests that bumblebees engage in featural generalization of the visual display of the string connected to the flower in this task. If the view of the string connected to the flower was restricted during the training phase, the proportion of bumblebees choosing the connected strings significantly decreased. Similarly, when the bumblebees were confronted with coiled connected strings during the testing phase, they failed to identify and reject the interrupted strings. This finding underscores the significance of visual consistency in enabling the bumblebees to perform the task successfully. Our results suggest that bumblebees' ability to distinguish between continuous strings and interrupted strings relies on a combination of image matching and associative learning, rather than means-end understanding. These insights contribute to a deeper understanding of the cognitive processes employed by bumblebees when tackling complex spatial tasks.

Asunto(s)

RESUMEN

Goal: In recent years, deep neural networks have consistently outperformed previously proposed methods in the domain of medical segmentation. However, due to their nature, these networks often struggle to delineate desired structures in data that fall outside their training distribution. The goal of this study is to address the challenges associated with domain generalization in CT segmentation by introducing a novel method called BucketAugment for deep neural networks. Methods: BucketAugment leverages principles from the Q-learning algorithm and employs validation loss to search for an optimal policy within a search space comprised of distributed stacks of 3D volumetric augmentations, termed 'buckets.' These buckets have tunable parameters and can be seamlessly integrated into existing neural network architectures, offering flexibility for customization. Results: In our experiments, we focus on segmenting kidney and liver structures across three distinct medical datasets, each containing CT scans of the abdominal region collected from various clinical institutions and scanner vendors. Our results indicate that BucketAugment significantly enhances domain generalization across diverse medical datasets, requiring only minimal modifications to existing network architectures. Conclusions: The introduction of BucketAugment provides a promising solution to the challenges of domain generalization in CT segmentation. By leveraging Q-learning principles and distributed stacks of 3D augmentations, this method improves the performance of deep neural networks on medical segmentation tasks, demonstrating its potential to enhance the applicability of such models across different datasets and clinical scenarios.

RESUMEN

We investigated autistic children's generalisation of social communication over time across three settings during a play-based assessment with different adults and explore the potential moderating effects on generalisation of age, nonverbal IQ and level of restricted and repetitive behaviours. The social communication abilities of 248 autistic children (2-11 years, 21% female, 22% single parent, 60% white) from three UK sites were assessed from 1984 video interactions in three contexts with three different interaction partners (parent/home, teaching assistant/school, researcher/clinic) at baseline, midpoint (+ 7m) and endpoint (+ 12m) within the Paediatric Autism Communication Trial-Generalised (PACT-G), a parent-mediated social communication intervention. Children's midpoint social communication at home generalised to school at midpoint and to clinic at endpoint. Generalisation was stronger from home to school and clinic than school to home and clinic. Generalisation was not moderated by age, nonverbal IQ or restricted and repetitive behaviour. Broader child development did not explain the pattern of results. The current study is the largest study to date to explore generalisation with autistic children and provides novel insight into their generalisation of social communication skills. Further research is needed to gain a more comprehensive understanding of facilitators of generalisation across settings and interaction partners in order to develop targeted strategies for interventions to enhance outcomes for young autistic children.

RESUMEN

Semantic Dementia (SD) is a neurodegenerative disease characterised by progressive deterioration of semantic knowledge, resulting in diminished understanding of concepts, whether encountered in verbal or non-verbal form. Over the past three decades, a number of studies employing a range of treatment techniques and learning methods have examined whether patients with SD can relearn previously known concepts or learn and retain new information. In this article, we review this research, addressing two main questions: a) Can aspects of semantic knowledge that are 'lost' due to degeneration be re-acquired? b) How much do other memory systems (working and episodic memory) interact with and depend on semantic memory? Several studies demonstrate successful relearning of previously known words and concepts in SD, particularly after regular, prolonged practice; but this success tends to diminish once practice ceases, and furthermore often fails to generalise to other instances of the same object/concept. This pattern suggests that, with impaired semantic knowledge, learning relies to an abnormal extent on perceptual factors, making it difficult to abstract away from the specific visual or other perceptual format in which a given concept has been trained. Furthermore, the impact of semantic 'status' of a word or object on both working and episodic memory indicates pervasive interaction of these other memory systems with conceptual knowledge.

Asunto(s)

RESUMEN

Fear and anxiety are rarely confined to specific stimuli or situations. In fear generalisation, there is a spread of fear responses elicited by physically dissimilar generalisation stimuli (GS) along a continuum between danger and safety. The current study investigated fear generalisation with a novel online task using COVID-19-relevant stimuli (i.e., busy or quiet shopping street/mall scenes) during pandemic lockdown restrictions in the United Kingdom. Participants (N = 50) first completed clinically relevant trait measures before commencing a habituation phase, where two conditioned stimuli (CSs; i.e., a busy or quiet high street/mall scene) were presented. Participants then underwent fear conditioning where one conditioned stimulus (CS+) was followed by an aversive unconditioned stimulus (US; a loud female scream accompanied by a facial photograph of a female displaying a fearful emotion) and another (CS-) was not. In a test phase, six generalisation stimuli were presented where the US was withheld, and participants provided threat expectancy and fear ratings for all stimuli. Following successful conditioning, fear generalization was observed for both threat expectancy and fear ratings. Trait worry partially predicted generalised threat expectancy and COVID-19 fear strongly predicted generalised fear. In conclusion, a generalisation gradient was evident using an online remote generalisation task with images of busy/quiet streets during the pandemic. Worry and fear of COVID-19 predicted fear generalisation.

Asunto(s)

RESUMEN

Previous research has found that episodic simulation of events of helping others can effectively enhance intentions to help the same person involved and the identical situational context as the imagined scenarios. This 'prosocial simulation effect' is argued to reflect, at least in part, associative memory mechanisms whereby the simulation is reactivated when in the same situation as that imagined. However, to date, no study has examined systematically whether this 'prosocial simulation effect' can be transferred to response scenarios involving different people and/or situational contexts to the imagined scenarios, and if so, whether the degree of overlap with the imagined helping episode modulated the transfer effect. Across two experiments, we systematically varied the overlap of the simulated and response scenarios, both in terms of the persons in need and/or the situational contexts, and whether would influence the magnitude of prosocial simulation effect. Results from both experiments showed that the prosocial simulation effect can be transferred to response scenarios involving different people and situational contexts to the simulated scenarios. However, this finding was primarily driven by response scenarios that had a high degree of overlap to the simulated scenarios. The application of our findings to the practical implementation of simulation to promote prosociality in the real world is discussed.

Asunto(s)

RESUMEN

Ideomotor theory suggests that selecting a response is achieved by anticipating the consequences of that response. Evidence for this is the response-effect compatibility (REC) effect, that is, responding tends to be faster when the (anticipated) predictable consequences of a response (the action effects) are compatible rather than incompatible with the response. The present experiments investigated the extent to which the consequences must be exactly versus categorically predictable. According to the latter, an abstraction from particular instances to the categories of dimensional overlap might take place. For participants in one group of Experiment 1, left-hand and right-hand responses produced compatible or incompatible action effects in perfectly predictable positions to the left or right of fixation, and a standard REC effect was observed. For participants in another group of Experiment 1, as well as in Experiments 2 and 3, the responses also produced action effects to the left or right of fixation, but the eccentricity of the action effects (and thus their precise location) was unpredictable. On average, the data from the latter groups suggest that there is little, if any, tendency for participants to abstract the critical left/right features from spatially somewhat unpredictable action effects and use them for action selection, although there were large individual differences in these groups. Thus, at least on average across participants, it appears that the spatial locations of action effects must be perfectly predictable for these effects to have a strong influence on the response time.

Asunto(s)

RESUMEN

PURPOSE: The purpose of the study was to investigate the treatment efficacy of a discourse-based working memory (WM) protocol for individuals with the amnestic type of mild cognitive impairment (MCI). METHOD: The current study employed a randomised, single-blind design. Fourteen individuals with MCI participated in the study (n = 7 treatment group and n = 7 control group). The treatment protocol consisted of 10 sessions two times per week, and treatment was individually administered only to the treatment group. A Wilcoxon signed-rank test was performed to verify pre-post comparisons within each group. Mann-Whitney nonparametric tests were conducted to confirm the differences between the treatment and control groups for the post-treatment scores. RESULT: The treatment group demonstrated a significant increase in story-retelling outcomes for both the treated stories and untreated novel stories compared to the control group. Furthermore, the treatment group presented transfer effects for WM span measures and controlled word association tasks. CONCLUSION: The results indicated that a discourse-based WM treatment protocol is efficacious for the amnestic type of mild cognitive impairment with the effects transferred to frontal lobe functions, as measured by WM tasks and semantic word fluency measures. Further studies are needed to track the trajectory of performance across sessions.

RESUMEN

Purpose: The conversion of ocular myasthenia gravis (OMG) to generalised myasthenia gravis (GMG) is reported to differ depending on the presence of generalisation risk factors (Mazzoli et al. 2018). Thymic pathology has been recognised as a potential risk factor for generalisation in the literature (Teo et al. 2017). Thymoma and thymic hyperplasia have yet to be examined as a risk factor for generalisation of OMG independently of other risk factors in the literature. Thus, the purpose of this review is to examine the literature to identify whether thymoma and thymic hyperplasia do increase the risk of OMG progressing to GMG. Methods: A literature search was carried out which employed a systematic approach. The search was undertaken using the following academic libraries: MEDLINE, Embase and Starplus. The search was limited to publications between the years 2001 to 2021. The search yielded 82 studies, which after the screening of titles and abstracts, left 62 studies for further analysis against the inclusion and exclusion criteria. Results: The review found thymoma to be associated with an increased risk of GMG development. However, there was a scarce amount of literature which investigated thymic hyperplasia. Therefore, a firm conclusion could not be made with regards to thymic hyperplasia and the risk of GMG development. Conclusions: This review provides evidence for the consideration of thymectomy early after thymomatous OMG diagnosis to prevent GMG conversion. As the review did not collect enough evidence to support the influence of thymic hyperplasia on OMG conversion, further research is required.

RESUMEN

We investigated whether learning an artificial language at 17 months was predictive of children's natural language vocabulary and grammar skills at 54 months. Children at 17 months listened to an artificial language containing non-adjacent dependencies, and were then tested on their learning to segment and to generalise the structure of the language. At 54 months, children were then tested on a range of standardised natural language tasks that assessed receptive and expressive vocabulary and grammar. A structural equation model demonstrated that learning the artificial language generalisation at 17 months predicted language abilities - a composite of vocabulary and grammar skills - at 54 months, whereas artificial language segmentation at 17 months did not predict language abilities at this age. Artificial language learning tasks - especially those that probe grammar learning - provide a valuable tool for uncovering the mechanisms driving children's early language development.

Asunto(s)

RESUMEN

We present a novel information-theoretic framework, termed as TURBO, designed to systematically analyse and generalise auto-encoding methods. We start by examining the principles of information bottleneck and bottleneck-based networks in the auto-encoding setting and identifying their inherent limitations, which become more prominent for data with multiple relevant, physics-related representations. The TURBO framework is then introduced, providing a comprehensive derivation of its core concept consisting of the maximisation of mutual information between various data representations expressed in two directions reflecting the information flows. We illustrate that numerous prevalent neural network models are encompassed within this framework. The paper underscores the insufficiency of the information bottleneck concept in elucidating all such models, thereby establishing TURBO as a preferable theoretical reference. The introduction of TURBO contributes to a richer understanding of data representation and the structure of neural network models, enabling more efficient and versatile applications.

RESUMEN

Category variability or diversity is an important factor influencing generalisation. However, expectations of category variability may not only depend on the variability of encountered category members, but may also be shaped by prior experiences with similar categories. In this study, we investigated whether we could influence category generalisation by inducing different category representations in an A/Non-A categorisation task: Participants either learned about a homogeneous category Non-A or a diverse category Non-A during a priming phase. To better understand the transfer process, we varied the nature of the learning phase from implicit transfer to explicit instructions that actively requested participants to use their prior experiences. We found that while with a homogeneous Non-A representation, generalisation of the A and Non-A categories was equal, the generalisation of category Non-A widened after a priming phase with a diverse representation. In a second experiment, we found that the widening of generalisation of category Non-A occurred when the exemplars in this category were themselves diverse (feature-diverse condition) but not when the category contained distinct exemplars (exemplar-diverse condition). These results suggests that categorisation is influenced by previous categorisation experiences possibly altering the representation of a category. Furthermore, the study gives a hint what kind of heterogeneity is needed to observe the commonly reported broader generalisation of diverse categories. The finding has implications not only to understand the influence of prior experiences on category learning, but any cognitive process that hinges on generalisation.

RESUMEN

BACKGROUND: Synthetic computed tomography (sCT) has been proposed and increasingly clinically adopted to enable magnetic resonance imaging (MRI)-based radiotherapy. Deep learning (DL) has recently demonstrated the ability to generate accurate sCT from fixed MRI acquisitions. However, MRI protocols may change over time or differ between centres resulting in low-quality sCT due to poor model generalisation. PURPOSE: investigating domain randomisation (DR) to increase the generalisation of a DL model for brain sCT generation. METHODS: CT and corresponding T1-weighted MRI with/without contrast, T2-weighted, and FLAIR MRI from 95 patients undergoing RT were collected, considering FLAIR the unseen sequence where to investigate generalisation. A "Baseline" generative adversarial network was trained with/without the FLAIR sequence to test how a model performs without DR. Image similarity and accuracy of sCT-based dose plans were assessed against CT to select the best-performing DR approach against the Baseline. RESULTS: The Baseline model had the poorest performance on FLAIR, with mean absolute error (MAE) = 106 ± 20.7 HU (mean ±σ). Performance on FLAIR significantly improved for the DR model with MAE = 99.0 ± 14.9 HU, but still inferior to the performance of the Baseline+FLAIR model (MAE = 72.6 ± 10.1 HU). Similarly, an improvement in γ-pass rate was obtained for DR vs Baseline. CONCLUSION: DR improved image similarity and dose accuracy on the unseen sequence compared to training only on acquired MRI. DR makes the model more robust, reducing the need for re-training when applying a model on sequences unseen and unavailable for retraining.

Asunto(s)

RESUMEN

Virtual reality (VR) is a promising tool for training life skills in people with intellectual disabilities. However, there is a lack of evidence surrounding the implementation, suitability, and effectiveness of VR training in this population. The present study investigated the effectiveness of VR training for people with intellectual disabilities by assessing (1) their ability to complete basic tasks in VR, (2) real-world transfer and skill generalisation, and (3) the individual characteristics of participants able to benefit from VR training. Thirty-two participants with an intellectual disability of varying severity completed a waste management training intervention in VR that involved sorting 18 items into three bins. Real-world performance was measured at pre-test, post-test, and delayed time points. The number of VR training sessions varied as training ceased when participants met the learning target (≈ 90% correct). A survival analysis assessed training success probability as a function of the number of training sessions with participants split by their level of adaptive functioning (as measured on the Adaptive Behaviour Assessment System Third Edition). The learning target was met by 19 participants (59.4%) within ten sessions (Mdn = 8.5, IQR 4-10). Real-world performance significantly improved from pre- to post-test and pre- to delayed test. There was no significant difference from post- to delayed test. Further, there was a significant positive relationship between adaptive functioning and change in the real-world assessment from the pre-test to the post- and delayed tests. VR facilitated the learning of most participants, which led to demonstrations of real-world transfer and skill generalisation. The present study identified a relationship between adaptive functioning and success in VR training. The survival curve may assist in planning future studies and training programs.

RESUMEN

Recent studies have shown that extinction training including the conditional stimulus (CS+) and stimuli that are similar to the CS + enhances extinction retention and generalisation to novel stimuli. However, in a clinical setting, the CS+ is rarely available for use during exposure therapy. The aim of the present study was to determine if replacing the CS+ with a similar versus different stimulus, and including other similar stimuli during extinction, could reduce fear at test on par with extinction using the original CS+ with and without other similar stimuli. In an experiment conducted in a single session, participants completed a habituation phase followed by an acquisition phase using two dog images presented with (CS+) and without (CS-) an acoustic unconditional stimulus (US). Participants were randomly allocated to four extinction conditions: similar CS + dog with novel dog images (Similar replacement extinction condition); different CS + dog with novel dog images (Different replacement extinction condition); original CS + dog with novel dog images (Multiple extinction control condition); and original CS + without novel dog images (Standard extinction control condition). All participants completed a test phase with the original CSs followed by a generalisation test with another two novel dog images. All groups acquired, and then extinguished differential skin conductance responses (SCRs) with no differences observed between groups. Whereas the Similar replacement extinction group and the Multiple and Standard extinction control groups did not exhibit significant differential SCRs when re-exposed to the original CS + relative to the CS- at test, differential responding to the CSs was significant at test in the Different replacement extinction group. There were no significant differences between groups in SCRs to the two novel dog images during the generalisation phase and in between-phase subjective ratings. Findings suggest that replacement stimuli used during extinction should be as similar as possible to the CS + to reduce physiological arousal to the original CS+.

Asunto(s)

RESUMEN

Overgeneralisation is one of the aetiologies of anxiety disorders and is closely associated with elevated intolerance of uncertainty (IU) levels. However, the underlying mechanisms are unclear. Considering the inconsistency of previous results and the high sensitivity of IU to uncertainty, the present study investigated the effect of IU on threat generalisation in predictable and unpredictable conditions. We compared self-reported unconditioned stimuli (US) expectancy and event-related potentials (ERPs) during generalisation in high IU (n = 34) and low IU (n = 35) participants. The results indicated that high IU was associated with higher US expectancy for generalisation stimuli (GS) than with low IU. At the electrophysiological level, compared to low IU, high IU showed increased P1 to ambiguous GS as well as decreased early late positive potential (LPP) to GS in unpredictable conditions, and no differential response to GS in late LPP in predictable conditions. These findings suggest that IU enhances threat generalisation and may be related to increased early automatic attention to ambiguous stimulus and inadequate late elaborate processing in a high uncertainty context. These findings might contribute to the treatment of mood disorders characterized by high IU.

Asunto(s)

RESUMEN

Threat generalisation is an adaptive process that is essential for individual survival. In contrast, over-generalisation is a pathological process that underlies the development of anxiety disorders. Intolerance of uncertainty (IU) is an individual characteristic known to influence threat generalisation by altering the responses to threat in uncertain situations. However, how it affects contextual threat generalisation remains unclear. Here, we used a novel paradigm to investigate whether contextual threat generalisation varied between individuals depending on their IU level (high or low) and the predictability of a situation (predictable or unpredictable). We analysed shock expectancy in 82 participants (age: 18-27 years) during threat acquisition and generalisation. Results showed that compared with the low IU group, the high IU group exhibited increased contextual threat generalisation to threat-related cues in unpredictable situation. These findings suggest that IU could be a marker for anxiety disorder susceptibility, as well as a target for anxiety treatment.

Asunto(s)

RESUMEN

Generalisation in evaluative conditioning occurs when the valence acquired by a conditional stimulus (CS), after repeated pairing with an unconditional stimulus (US), spreads to stimuli that are similar to the CS (generalisation stimuli, GS). CS evaluations can be updated via CS instructions that conflict with prior conditioning (negative conditioning + positive instruction). We examined whether CS instructions can update GS evaluations after conditioning. We used alien stimuli where one alien (CSp) from a fictional group was paired with pleasant US images and another alien (CSu) from a different group was paired with unpleasant US images. The other members from the two groups were used as GSs. After conditioning, participants received negative CSp instructions and positive CSu instructions. In Experiment 1, explicit and implicit GS evaluations were measured before and after the instructions. In Experiment 2, we used a between-participants design where one group received positive/negative CS instructions while a control group received neutral instructions. In both experiments, the positive/negative CS instructions caused a reversal of explicit GS evaluations and an elimination of implicit GS evaluations. The findings suggest that generalised evaluations can change after CS instructions which may have implications for interventions aimed at reducing negative group attitudes.

Asunto(s)

RESUMEN

Given the prevalence of depression worldwide and its major impact on society, several studies employed artificial intelligence modelling to automatically detect and assess depression. However, interpretation of these models and cues are rarely discussed in detail in the AI community, but have received increased attention lately. In this study, we aim to analyse the commonly selected features using a proposed framework of several feature selection methods and their effect on the classification results, which will provide an interpretation of the depression detection model. The developed framework aggregates and selects the most promising features for modelling depression detection from 38 feature selection algorithms of different categories. Using three real-world depression datasets, 902 behavioural cues were extracted from speech behaviour, speech prosody, eye movement and head pose. To verify the generalisability of the proposed framework, we applied the entire process to depression datasets individually and when combined. The results from the proposed framework showed that speech behaviour features (e.g. pauses) are the most distinctive features of the depression detection model. From the speech prosody modality, the strongest feature groups were F0, HNR, formants, and MFCC, while for the eye activity modality they were left-right eye movement and gaze direction, and for the head modality it was yaw head movement. Modelling depression detection using the selected features (even though there are only 9 features) outperformed using all features in all the individual and combined datasets. Our feature selection framework did not only provide an interpretation of the model, but was also able to produce a higher accuracy of depression detection with a small number of features in varied datasets. This could help to reduce the processing time needed to extract features and creating the model.

RESUMEN

The aim of the current study was to conduct a reliability generalisation (RG) meta-analysis of Cronbach's alpha for the Insomnia Severity Index (ISI). A systematic search of three databases (PubMed, Scopus, and Web of Science) from inception to 12 March 2021 was performed. Publications that reported Cronbach's alpha for the total ISI score were included. Only psychometric-focussed studies were considered. Meta-analysis was carried out using a random-effects model to derive a pooled estimate of Cronbach's alphas. The number of participants in the included publications ranged from 25 to 12,056, with 33 studies (42 estimates) comprising internal consistency coefficients, and a combined sample size of N = 29,688. The age range of the included publications was from 13.4 to 74.3 years. Data extraction implied 33 publications out of 706 found through the database search. Cronbach's alphas ranged from 0.65 to 0.92. The majority of the reported coefficients were ≥0.7 and presented a low risk of bias (n = 32). The pooled alpha coefficient was 0.83 (IC [0.81-0.85]; SE = 0.009) with high heterogeneity among the included publications (I2  = 97%). Subgroup analyses including moderators such as continent, setting, risk of bias, and age did not affect significantly the overall result. In general, the cumulative estimate of Cronbach's alpha for the ISI is good. However, this finding should be interpreted with caution since there is a high heterogeneity level and some of the studies might not have checked the assumptions underlying Cronbach's alphas.

Asunto(s)