Development and application of CatFACS: Are human cat adopters influenced by cat facial expressions?

From facial expressions to algorithms: a narrative review of animal pain recognition technologies

Facial expressions are essential for communication and emotional expression across species. Despite the improvements brought by tools like the Horse Grimace Scale (HGS) in pain recognition in horses, their reliance on human identification of characteristic traits presents drawbacks such as subjectivity, training requirements, costs, and potential bias. Despite these challenges, the development of facial expression pain scales for animals has been making strides. To address these limitations, Automated Pain Recognition (APR) powered by Artificial Intelligence (AI) offers a promising advancement. Notably, computer vision and machine learning have revolutionized our approach to identifying and addressing pain in non-verbal patients, including animals, with profound implications for both veterinary medicine and animal welfare. By leveraging the capabilities of AI algorithms, we can construct sophisticated models capable of analyzing diverse data inputs, encompassing not only facial expressions but also body language, vocalizations, and physiological signals, to provide precise and objective evaluations of an animal's pain levels. While the advancement of APR holds great promise for improving animal welfare by enabling better pain management, it also brings forth the need to overcome data limitations, ensure ethical practices, and develop robust ground truth measures. This narrative review aimed to provide a comprehensive overview, tracing the journey from the initial application of facial expression recognition for the development of pain scales in animals to the recent application, evolution, and limitations of APR, thereby contributing to understanding this rapidly evolving field.

Fully automated deep learning models with smartphone applicability for prediction of pain using the Feline Grimace Scale

This study used deep neural networks and machine learning models to predict facial landmark positions and pain scores using the Feline Grimace Scale© (FGS). A total of 3447 face images of cats were annotated with 37 landmarks. Convolutional neural networks (CNN) were trained and selected according to size, prediction time, predictive performance (normalized root mean squared error, NRMSE) and suitability for smartphone technology. Geometric descriptors (n = 35) were computed. XGBoost models were trained and selected according to predictive performance (accuracy; mean square error, MSE). For prediction of facial landmarks, the best CNN model had NRMSE of 16.76% (ShuffleNetV2). For prediction of FGS scores, the best XGBoost model had accuracy of 95.5% and MSE of 0.0096. Models showed excellent predictive performance and accuracy to discriminate painful and non-painful cats. This technology can now be used for the development of an automated, smartphone application for acute pain assessment in cats.

Feline faces: Unraveling the social function of domestic cat facial signals

Lately, there has been a growing interest in studying domestic cat facial signals, but most of this research has centered on signals produced during human-cat interactions or pain. The available research on intraspecific facial signaling with domesticated cats has largely focused on non-affiliative social interactions. However, the transition to intraspecific sociality through domestication could have resulted in a greater reliance on affiliative facial signals that aid with social bonding. Our study aimed to document the various facial signals that cats produce during affiliative and non-affiliative intraspecific interactions. Given the close relationship between the physical form and social function of mammalian facial signals, we predicted that affiliative and non-affiliative facial signals would have noticeable differences in their physical morphology. We observed the behavior of 53 adult domestic shorthair cats at CatCafé Lounge in Los Angeles, CA. Using Facial Action Coding Systems designed for cats, we compared the complexity and compositionality of facial signals produced in affiliative and non-affiliative contexts. To measure complexity and compositionality, we examined the number and types of facial muscle movements (AUs) observed in each signal. We found that compositionality, rather than complexity, was significantly associated with the social function of intraspecific facial signals. Our findings indicate that domestication likely had a significant impact on the development of intraspecific facial signaling repertoires in cats.

Explainable automated pain recognition in cats

Manual tools for pain assessment from facial expressions have been suggested and validated for several animal species. However, facial expression analysis performed by humans is prone to subjectivity and bias, and in many cases also requires special expertise and training. This has led to an increasing body of work on automated pain recognition, which has been addressed for several species, including cats. Even for experts, cats are a notoriously challenging species for pain assessment. A previous study compared two approaches to automated 'pain'/'no pain' classification from cat facial images: a deep learning approach, and an approach based on manually annotated geometric landmarks, reaching comparable accuracy results. However, the study included a very homogeneous dataset of cats and thus further research to study generalizability of pain recognition to more realistic settings is required. This study addresses the question of whether AI models can classify 'pain'/'no pain' in cats in a more realistic (multi-breed, multi-sex) setting using a more heterogeneous and thus potentially 'noisy' dataset of 84 client-owned cats. Cats were a convenience sample presented to the Department of Small Animal Medicine and Surgery of the University of Veterinary Medicine Hannover and included individuals of different breeds, ages, sex, and with varying medical conditions/medical histories. Cats were scored by veterinary experts using the Glasgow composite measure pain scale in combination with the well-documented and comprehensive clinical history of those patients; the scoring was then used for training AI models using two different approaches. We show that in this context the landmark-based approach performs better, reaching accuracy above 77% in pain detection as opposed to only above 65% reached by the deep learning approach. Furthermore, we investigated the explainability of such machine recognition in terms of identifying facial features that are important for the machine, revealing that the region of nose and mouth seems more important for machine pain classification, while the region of ears is less important, with these findings being consistent across the models and techniques studied here.

Explainable automated recognition of emotional states from canine facial expressions: the case of positive anticipation and frustration

In animal research, automation of affective states recognition has so far mainly addressed pain in a few species. Emotional states remain uncharted territories, especially in dogs, due to the complexity of their facial morphology and expressions. This study contributes to fill this gap in two aspects. First, it is the first to address dog emotional states using a dataset obtained in a controlled experimental setting, including videos from (n = 29) Labrador Retrievers assumed to be in two experimentally induced emotional states: negative (frustration) and positive (anticipation). The dogs' facial expressions were measured using the Dogs Facial Action Coding System (DogFACS). Two different approaches are compared in relation to our aim: (1) a DogFACS-based approach with a two-step pipeline consisting of (i) a DogFACS variable detector and (ii) a positive/negative state Decision Tree classifier; (2) An approach using deep learning techniques with no intermediate representation. The approaches reach accuracy of above 71% and 89%, respectively, with the deep learning approach performing better. Secondly, this study is also the first to study explainability of AI models in the context of emotion in animals. The DogFACS-based approach provides decision trees, that is a mathematical representation which reflects previous findings by human experts in relation to certain facial expressions (DogFACS variables) being correlates of specific emotional states. The deep learning approach offers a different, visual form of explainability in the form of heatmaps reflecting regions of focus of the network's attention, which in some cases show focus clearly related to the nature of particular DogFACS variables. These heatmaps may hold the key to novel insights on the sensitivity of the network to nuanced pixel patterns reflecting information invisible to the human eye.

Going Deeper than Tracking: A Survey of Computer-Vision Based Recognition of Animal Pain and Emotions

Advances in animal motion tracking and pose recognition have been a game changer in the study of animal behavior. Recently, an increasing number of works go ‘deeper’ than tracking, and address automated recognition of animals’ internal states such as emotions and pain with the aim of improving animal welfare, making this a timely moment for a systematization of the field. This paper provides a comprehensive survey of computer vision-based research on recognition of pain and emotional states in animals, addressing both facial and bodily behavior analysis. We summarize the efforts that have been presented so far within this topic—classifying them across different dimensions, highlight challenges and research gaps, and provide best practice recommendations for advancing the field, and some future directions for research.

Revisiting Darwin's comparisons between human and non-human primate facial signals

Darwin and other pioneering scholars made comparisons between human facial signals and those of non-human primates, suggesting they share evolutionary history. We now have tools available (Facial Action Coding System: FACS) to make these comparisons anatomically based and standardised, as well as analytical methods to facilitate comparative studies. Here we review the evidence establishing a shared anatomical basis between the facial behaviour of human and non-human primate species, concluding which signals are likely related, and which are not. We then review the evidence for shared function and discuss the implications for understanding human communication. Where differences between humans and other species exist, we explore possible explanations and future directions for enquiry.

Automated recognition of pain in cats

Facial expressions in non-human animals are closely linked to their internal affective states, with the majority of empirical work focusing on facial shape changes associated with pain. However, existing tools for facial expression analysis are prone to human subjectivity and bias, and in many cases also require special expertise and training. This paper presents the first comparative study of two different paths towards automatizing pain recognition in facial images of domestic short haired cats (n = 29), captured during ovariohysterectomy at different time points corresponding to varying intensities of pain. One approach is based on convolutional neural networks (ResNet50), while the other-on machine learning models based on geometric landmarks analysis inspired by species specific Facial Action Coding Systems (i.e. catFACS). Both types of approaches reach comparable accuracy of above 72%, indicating their potential usefulness as a basis for automating cat pain detection from images.

The Effect of Baby Schema in Cats on Length of Stay in an Irish Animal Shelter

Simple Summary

Several factors may influence cats’ length of stay (LoS) within the shelter environment. The aim of this study was to investigate the potential influence that ‘cuteness’ has on cats’ length of stay (LoS) in an animal shelter. Other factors such as paired-homing requirements, adoption profile (advertisements to encourage the adoption of a specific animal) wording, coat colour, sex and age were also explored. ‘Cuteness’ was measured in two ways; 1. Facial characteristics which characterise human infants, such as big eyes and a round head, were measured from images of cats to create an objective cuteness score. 2. Subjective cuteness scores were developed by participants rating the ‘cuteness’ of the cat images from ‘1-Not very cute’ to ‘5-Extremely cute’ through two online surveys; survey 1 investigated the cats’ social nature as well as cuteness while the second only looked at cuteness. The subjective cuteness scores were used to validate the objective cuteness scores. The analysis found that subjective cuteness influenced cats’ length of stay during the analysis of the first survey with no effect of other variables such as adoption profile language, homing requirements, coat colour, sex and age. Analysis of the data from the second survey found that none of the variables influenced the cats’ LoS. A novel objective measurement for cuteness in cats was developed which reflects the shape of the eyes and is associated with subjective cuteness scores. This research looks to understand factors that influence LoS to help identify potential methods to reduce time spent in shelters and improve the welfare of cats within these environments.

Abstract

Several factors may influence cats’ length of stay (LoS) within the shelter environment. The aim of this study was to investigate the potential influence that baby schema (characteristic facial features of infants, related to perceived ‘cuteness’) has on cats’ length of stay (LoS) in an animal shelter. Other factors such as paired-homing requirements, adoption profile (adverts to encourage the adoption of a specific animal) wording, coat colour, sex and age were also explored. Objective ‘cuteness’ scores were created by measuring specific facial features of 165 shelter cats. Several proportions of the cats’ faces were explored to identify the best objective measure of cuteness, including elements which have been found to associate with cuteness in cats and humans and new exploratory eye measurements. Subjective cuteness scores were developed by participants rating the ‘cuteness’ of the cat from ‘1-Not very cute’ to ‘5-Extremely cute’ through two online surveys; survey 1 investigated the cats’ social nature as well as cuteness while the second only looked at cuteness. The subjective cuteness scores were used to validate the objective cuteness scores. The analysis found that subjective cuteness in survey 1 was the only variable of influence on LoS. A novel objective cuteness measurement was developed which reflects the eye shape associated with subjective cuteness scores. The current study found that cuteness may not be as influential on cats’ LoS in shelters as hypothesised. This research looks to understand factors which influence LoS to help identify potential methods to reduce time spent in shelters and improve the welfare of cats within these environments.

CalliFACS: The common marmoset Facial Action Coding System

Facial expressions are subtle cues, central for communication and conveying emotions in mammals. Traditionally, facial expressions have been classified as a whole (e.g. happy, angry, bared-teeth), due to automatic face processing in the human brain, i.e., humans categorise emotions globally, but are not aware of subtle or isolated cues such as an eyebrow raise. Moreover, the same facial configuration (e.g. lip corners pulled backwards exposing teeth) can convey widely different information depending on the species (e.g. humans: happiness; chimpanzees: fear). The Facial Action Coding System (FACS) is considered the gold standard for investigating human facial behaviour and avoids subjective interpretations of meaning by objectively measuring independent movements linked to facial muscles, called Action Units (AUs). Following a similar methodology, we developed the CalliFACS for the common marmoset. First, we determined the facial muscular plan of the common marmoset by examining dissections from the literature. Second, we recorded common marmosets in a variety of contexts (e.g. grooming, feeding, play, human interaction, veterinary procedures), and selected clips from online databases (e.g. YouTube) to identify their facial movements. Individual facial movements were classified according to appearance changes produced by the corresponding underlying musculature. A diverse repertoire of 33 facial movements was identified in the common marmoset (15 Action Units, 15 Action Descriptors and 3 Ear Action Descriptors). Although we observed a reduced range of facial movement when compared to the HumanFACS, the common marmoset's range of facial movements was larger than predicted according to their socio-ecology and facial morphology, which indicates their importance for social interactions. CalliFACS is a scientific tool to measure facial movements, and thus, allows us to better understand the common marmoset's expressions and communication. As common marmosets have become increasingly popular laboratory animal models, from neuroscience to cognition, CalliFACS can be used as an important tool to evaluate their welfare, particularly in captivity.

Current Advances in Assessment of Dog's Emotions, Facial Expressions, and Their Use for Clinical Recognition of Pain

Animals' facial expressions are involuntary responses that serve to communicate the emotions that individuals feel. Due to their close co-existence with humans, broad attention has been given to identifying these expressions in certain species, especially dogs. This review aims to analyze and discuss the advances in identifying the facial expressions of domestic dogs and their clinical utility in recognizing pain as a method to improve daily practice and, in an accessible and effective way, assess the health outcome of dogs. This study focuses on aspects related to the anatomy and physiology of facial expressions in dogs, their emotions, and evaluations of their eyebrows, eyes, lips, and ear positions as changes that reflect pain or nociception. In this regard, research has found that dogs have anatomical configurations that allow them to generate changes in their expressions that similar canids-wolves, for example-cannot produce. Additionally, dogs can perceive emotions similar to those of their human tutors due to close human-animal interaction. This phenomenon-called "emotional contagion"-is triggered precisely by the dog's capacity to identify their owners' gestures and then react by emitting responses with either similar or opposed expressions that correspond to positive or negative stimuli, respectively. In conclusion, facial expressions are essential to maintaining social interaction between dogs and other species, as in their bond with humans. Moreover, this provides valuable information on emotions and the perception of pain, so in dogs, they can serve as valuable elements for recognizing and evaluating pain in clinical settings.

NetFACS: Using network science to understand facial communication systems

Understanding facial signals in humans and other species is crucial for understanding the evolution, complexity, and function of the face as a communication tool. The Facial Action Coding System (FACS) enables researchers to measure facial movements accurately, but we currently lack tools to reliably analyse data and efficiently communicate results. Network analysis can provide a way to use the information encoded in FACS datasets: by treating individual AUs (the smallest units of facial movements) as nodes in a network and their co-occurrence as connections, we can analyse and visualise differences in the use of combinations of AUs in different conditions. Here, we present ‘NetFACS’, a statistical package that uses occurrence probabilities and resampling methods to answer questions about the use of AUs, AU combinations, and the facial communication system as a whole in humans and non-human animals. Using highly stereotyped facial signals as an example, we illustrate some of the current functionalities of NetFACS. We show that very few AUs are specific to certain stereotypical contexts; that AUs are not used independently from each other; that graph-level properties of stereotypical signals differ; and that clusters of AUs allow us to reconstruct facial signals, even when blind to the underlying conditions. The flexibility and widespread use of network analysis allows us to move away from studying facial signals as stereotyped expressions, and towards a dynamic and differentiated approach to facial communication.

Can a facial action coding system (CatFACS) be used to determine the welfare state of cats with cerebellar hypoplasia?

BACKGROUND

The impaired motor skills of cats living with cerebellar hypoplasia (CH) suggests they would be unable to practice normal behaviour, one of the five welfare needs. This study aimed to explore the use of facial action coding system (CatFACS) as a welfare assessment tool for cats with CH.

METHODS

Facial expressions (action units [AUs]) were defined as neutral/positive or negative by recording healthy cats (n = 89) during presumed aversive or relaxed scenarios. CH cats (n = 33) were then filmed and their facial expressions compared to those of the presumed positively- and negatively-valenced healthy cats.

RESULTS

Sixteen negative AUs were defined. CH cats performed more of these than healthy cats (p = 0.023) in the relaxed scenario. There was no difference in AU expression between three levels of CH severity (mild, moderate or severe) (p = 0.461).

CONCLUSION

Cats perform distinct AUs when experiencing negatively-valenced arousal, the presence or absence of these AUs could be used to infer the welfare of healthy and CH cats. As there was no difference in AU expression between the three levels of CH severity, the behavioural restrictions CH imposes on cats does not necessarily indicate lower welfare and the reasons why CH cats perform more negatively associated AUs warrant further research.

Heads and Tails: An Analysis of Visual Signals in Cats, Felis catus

Simple Summary

Communication between individuals of the same species is essential in their interactions to regulate their proximities and distances. Communication includes exchanges of more or less complex visual signals. We attempt to decipher the most significant features of a visual configuration involving the combination of tail and ear positions in interactions between cats. Although the tail is a conspicuous feature for human observers, we demonstrate that ear positions of the cats in dyadic interactions with other cats are the best predictor of the outcomes of these interactions. However, in cat–human interactions, the cat most often approached with its tail up prior to rubbing itself against the human. The results are important for a better understanding of cats’ perceptions of humans, and will help to promote cat welfare.

Abstract

Visual communication involves specific signals. These include the different positions of mobile body elements. We analyzed visual configurations in cats that involve ears and the tail. We aimed at deciphering which features of these configurations were the most important in cats’ interactions with other cats and with humans. We observed a total of 254 cat–cat interactions within a sample of 29 cats, during a total of 100 h of observation scheduled with the “Behavioral dependent onset of sampling” method and using the “All occurences” sampling method. In addition, we sampled 10 interactions between cats and humans. In cat–cat interactions, we noted the positions of ears and tail of both protagonists, as well as the outcome of the interaction, which was either positive/neutral or negative. In a great majority of the 254 interactions sampled, both cats held their tail down. On the contrary, ear position was a critical element in predicting the outcome. When both partners held their ears erect, the outcome was significantly positive, such as rubbing or close proximity. In all other cases of the position of ears in both cats, the outcome was negative, with increased distance of the partners. Although the tail did not seem to play a significant role in visual configurations in cat interactions, the “tail-up” display was important when a cat approached a human being. In the vast majority of cases the cat rubbed itself on a human’s leg(s). Thus, we may conclude that the presence of a human has a specific meaning in the cat’s world, probably as the result of a long period of commensalism. It is important for pet owners to understand the signals that cats use with other cats and with humans in order to promote the welfare of cats.

Providing Humans With Practical, Best Practice Handling Guidelines During Human-Cat Interactions Increases Cats' Affiliative Behaviour and Reduces Aggression and Signs of Conflict

The importance of animals' experiences and associated comfort during Human-Animal Interactions (HAI), and particularly Animal Assisted Interventions (AAI), are increasingly recognised. However, there remains a paucity of published research, particularly concerning less formal but frequent HAIs to which companion animals are typically exposed, such as stroking or petting. Additionally, few practical evidence-based guides to facilitate humans' optimal animal handling and interaction in these contexts exist. A simple set of Human-Cat Interaction (HCI) guidelines were therefore created, with the aim to enhance domestic cats' comfort during generic HCI contexts. Based around a “CAT” acronym, guidelines focused on providing the cat with choice and control (“C”), paying attention (“A”) to the cats' behaviour and body language and limiting touch (“T”), primarily to their temporal regions. Guidelines were presented to human participants during a brief training intervention, and guideline efficacy was subsequently assessed. Domestic cats available for rehoming at Battersea Dogs and Cats Home, UK (n = 100) were filmed during interactions with novel members of the public (n = 120). Cats were exposed to a maximum of six, 5-min interaction sessions, balanced across “control” (interactions with humans pre-training) and “intervention” conditions (interactions with humans post-training). For each observation, cat behaviour and posture were coded and humans' cat-directed behaviour rated on the degree to which it reflected best practice principles. Data were extracted from a total of 535 observations and average human interaction ratings and cat behaviour values compared between control and intervention conditions via paired Wilcoxon tests. Compared to the control, humans' interaction styles were rated as significantly more closely aligned with best practice principles in the intervention condition. Cats also displayed significantly greater frequencies and/or durations of affiliative and positively-valenced behaviours in the intervention. In contrast, cats in the control displayed significantly greater frequencies of human-directed aggression, in addition to greater frequencies and/or durations of behaviours associated with conflict and negative valence. Results demonstrate the positive impact of practical interaction guidelines on cats' social behaviour and comfort during HCI, with the potential to improve cats' general experiences during interactions, reduce human-directed aggression and ultimately improve cat-human relationships.

Neural Responses of Pet Dogs Witnessing Their Caregiver's Positive Interactions with a Conspecific: An fMRI Study

We have limited knowledge on how dogs perceive humans and their actions. Various researchers investigated how they process human facial expressions, but their brain responses to complex social scenarios remain unclear. While undergoing fMRI, we exposed pet dogs to videos showing positive social and neutral nonsocial interactions between their caregivers and another conspecific. Our main interest was how the dogs responded to their caregivers (compared to a stranger) engaging in a pleasant interaction with another dog that could be seen as social rival. We hypothesized that the dogs would show activation increases in limbic areas such as the amygdala, hypothalamus, and insula and likely show higher attention and arousal during the positive caregiver-dog interaction. When contrasting the social with the nonsocial interaction, we found increased activations in the left amygdala and the insular cortex. Crucially, the dogs' hypothalamus showed strongest activation when the caregiver engaged in a positive social interaction. These findings indicate that dogs are sensitive to social affective human-dog interactions and likely show higher valence attribution and arousal in a situation possibly perceived as a potential threat to their caregiver bonds. Our study provides a first window into the neural correlates of social and emotional processing in dogs.

Using the Judgment Bias Task to Identify Behavioral Indicators of Affective State: Do Eye Wrinkles in Horses Reflect Mood?

Identifying and validating behavioral indicators of mood are important for the assessment of animal welfare. Here, we investigated whether horses' eye wrinkle expression in a presumably neutral situation is a measure of mood as assessed in a cognitive judgment bias task (JBT). To this end, we scored pictures of the left and right eyes of 16 stallions for different aspects of eye wrinkle expression and tested the same individuals on a spatial JBT with active trial initiation. Eye wrinkle expressions were assessed by a qualitative assessment, i.e., the overall assessment of how "worried" horses look, the number of wrinkles, and the angle measured at the intersection of lines drawn through the eyeball and the topmost wrinkle. Correlations between the three eye wrinkle measures and the optimism index as a measure of horses' decisions in the JBT were not statistically significant, but with increasing optimism index, horses tended to be scored as looking less worried (qualitative assessment). We discuss our findings from different perspectives and make suggestions for future research, e.g., by calling for experimental induction of mood and thus greater variation within and/or between individuals and by investigating the interplay between shorter-lasting emotional and longer-lasting mood states to further explore the potential use of the JBT to validate eye wrinkles and other facial or body expressions as indicators of mood.

Towards Machine Recognition of Facial Expressions of Pain in Horses

Automated recognition of human facial expressions of pain and emotions is to a certain degree a solved problem, using approaches based on computer vision and machine learning. However, the application of such methods to horses has proven difficult. Major barriers are the lack of sufficiently large, annotated databases for horses and difficulties in obtaining correct classifications of pain because horses are non-verbal. This review describes our work to overcome these barriers, using two different approaches. One involves the use of a manual, but relatively objective, classification system for facial activity (Facial Action Coding System), where data are analyzed for pain expressions after coding using machine learning principles. We have devised tools that can aid manual labeling by identifying the faces and facial keypoints of horses. This approach provides promising results in the automated recognition of facial action units from images. The second approach, recurrent neural network end-to-end learning, requires less extraction of features and representations from the video but instead depends on large volumes of video data with ground truth. Our preliminary results suggest clearly that dynamics are important for pain recognition and show that combinations of recurrent neural networks can classify experimental pain in a small number of horses better than human raters.

Bodily emotional expressions are a primary source of information for dogs, but not for humans

Dogs have remarkable abilities to synergise their behaviour with that of people, but how dogs read facial and bodily emotional cues in comparison to humans remains unclear. Both species share the same ecological niche, are highly social and expressive, making them an ideal comparative model for intra- and inter-species emotion perception. We compared eye-tracking data from unrestrained humans and dogs when viewing dynamic and naturalistic emotional expressions in humans and dogs. Dogs attended more to the body than the head of human and dog figures, unlike humans who focused more on the head of both species. Dogs and humans also showed a clear age effect that reduced head gaze. Our results indicate a species-specific evolutionary adaptation for emotion perception, which is only partly modified for heterospecific cues. These results have important implications for managing the risk associated with human-dog interactions, where expressive and perceptual differences are crucial.

Extending the MaqFACS to measure facial movement in Japanese macaques (Macaca fuscata) reveals a wide repertoire potential

Facial expressions are complex and subtle signals, central for communication and emotion in social mammals. Traditionally, facial expressions have been classified as a whole, disregarding small but relevant differences in displays. Even with the same morphological configuration different information can be conveyed depending on the species. Due to a hardwired processing of faces in the human brain, humans are quick to attribute emotion, but have difficulty in registering facial movement units. The well-known human FACS (Facial Action Coding System) is the gold standard for objectively measuring facial expressions, and can be adapted through anatomical investigation and functional homologies for cross-species systematic comparisons. Here we aimed at developing a FACS for Japanese macaques, following established FACS methodology: first, we considered the species' muscular facial plan; second, we ascertained functional homologies with other primate species; and finally, we categorised each independent facial movement into Action Units (AUs). Due to similarities in the rhesus and Japanese macaques' facial musculature, the MaqFACS (previously developed for rhesus macaques) was used as a basis to extend the FACS tool to Japanese macaques, while highlighting the morphological and appearance changes differences between the two species. We documented 19 AUs, 15 Action Descriptors (ADs) and 3 Ear Action Units (EAUs) in Japanese macaques, with all movements of MaqFACS found in Japanese macaques. New movements were also observed, indicating a slightly larger repertoire than in rhesus or Barbary macaques. Our work reported here of the MaqFACS extension for Japanese macaques, when used together with the MaqFACS, comprises a valuable objective tool for the systematic and standardised analysis of facial expressions in Japanese macaques. The MaqFACS extension for Japanese macaques will now allow the investigation of the evolution of communication and emotion in primates, as well as contribute to improving the welfare of individuals, particularly in captivity and laboratory settings.

The Application of Geometric Morphometrics to Explore Potential Impacts of Anthropocentric Selection on Animals' Ability to Communicate via the Face: The Domestic Cat as a Case Study

During their domestication via artificial selection, humans have substantially modified the morphology and thus visual appearance of non-human animals. While research highlights the negative impact of these modifications on physical functioning, little is known about their impact on behavior and signaling, either toward humans or conspecifics. Changes in the appearance of the face, such as those associated with, but not limited to, facial expressions, form an important part of non-verbal communication. In companion animals, the face is one of their most visually diverse features (due to human-driven selection), which may impact the visual clarity of expressions and other forms of signaling. Using the domestic cat as our model, we applied a new analytical technique in order to understand the impact of breed variation on relative positioning of facial landmarks, chosen specifically for their association with the production of various facial movements, and the expression of affect. We then assessed the extent to which facial appearances known to be associated with a specific underlying state (i.e., pain, assessed via a validated, facial pain score), could be reliably detected in a morphologically diverse population. Substantial baseline variation in landmarks was identified at both the cephalic (e.g., brachycephalic, dolichocephalic, mesocephalic) as well as breed levels. While differences in facial pain scores could successfully differentiate between "pain" and "no pain" in the facial appearance of domestic shorthaired cats (DSH), these differences were no longer detectable when assessed within a larger more morphologically diverse population, after corrections for multiple testing were applied. There was also considerable overlap between pain scores in the DSH "pain" population and the neutral faces of other breeds. Additionally, for several paedomorphic breeds, their neutral face shapes produced scores indicative of greater pain, compared to most other breeds, including the DSH cats actually in pain. Our findings highlight the degree to which anthropocentric selection might disrupt the communicative content of animals' faces, in this case the domestic cat. These results also suggest a potential human preference for features extending beyond the infantile, to include negatively-valenced facial forms such as pain.

Slow Blink Eye Closure in Shelter Cats is Related to Quicker Adoption

Simple Summary

Slow blinking is a type of interaction between humans and cats that involves a sequence of prolonged eye narrowing movements being given by both parties. This interspecific social behaviour has recently been studied empirically and appears to be a form of positive communication for cats, who are more likely to approach a previously unfamiliar human after such interactions. We investigated whether slow blinking can also affect human preferences for cats in a shelter environment. We measured whether cats’ readiness to respond to a human-initiated slow blink interaction was associated with rates of rehoming in the shelter. We also examined cats’ propensity to slow blink when they were anxious around humans or not. We demonstrated that cats that responded to human slow blinking by using eye closures themselves were rehomed quicker than cats that closed their eyes less. Cats that were initially identified as more nervous around humans also showed a trend towards giving longer total slow blink movements in response to human slow blinking. Our results suggest that the cat slow blink sequence is perceived as positive by humans and may have a dual function in cats, occurring in both affiliative and submissive situations.

Abstract

The process of domestication is likely to have led to the development of adaptive interspecific social abilities in animals. Such abilities are particularly interesting in less gregarious animals, such as cats. One notable social behaviour that cats exhibit in relation to humans is the slow blink sequence, which our previous research suggests can function as a form of positive communication between cats and humans. This behaviour involves the production of successive half blinks followed by either a prolonged narrowing of the eye or an eye closure. The present study investigates how cat (n = 18) slow blink sequences might affect human preferences during the adoption of shelter cats. Our study specifically tested (1) whether cats’ propensity to respond to human-initiated slow blinking was associated with their speed of rehoming from a shelter environment, and (2) whether cats’ anxiety around humans was related to their tendency to slow blink. Our experiments demonstrated that cats that showed an increased number of and longer eye closures in response to human slow blinks were rehomed faster, and that nervous cats, who had been identified as needing desensitisation to humans, tended to spend more time producing slow blink sequences in response to human slow blinks than a non-desensitisation group. Collectively, these results suggest that the cat slow blink sequence is perceived as positive by humans and may have a dual function—occurring in both affiliative and submissive contexts.

The Power of a Positive Human-Animal Relationship for Animal Welfare

Domestic animals often seek and enjoy interacting with humans. Positive human-animal relationships can elicit positive emotions and other positive welfare outcomes. Nevertheless, our understanding of the underlying processes that govern the positive perception of humans by animals is incomplete. We cover the potential mechanisms involved in the development and maintenance of positive human-animal relationships from the perspective of the animal. This encompasses habituation, associative learning, and possibly attachment or bonding based on communication and social cognition. We review the indicators from the literature to assess a positive human-animal relationship. We operationally define this positive relationship as the animal showing voluntary approach and spatial proximity (seeking) and signs of anticipation, pleasure, relaxation, or other indicators of a rewarding experience from interacting with the human. For research, we recommend accounting for the baseline human-animal relationship in the animal's everyday life, and incorporating a control treatment rather than only comparing positive to negative interaction treatments. Furthermore, animal characteristics, such as previous experience, genetics, and individual predisposition, as well as contextual characteristics related to the social and physical environment, may modulate the perception of humans by animals. The human-animal relationship is also influenced by human characteristics, such as the person's familiarity to the animal, attitudes, skills, and knowledge. We highlight implications for current practices and suggest simple solutions, such as paying attention to the animal's behavioral response to humans and providing choice and control to the animal in terms of when and how to interact with humans. Practical applications to achieve a positive perception of humans could be better utilized, such as by incorporating training principles, while keeping in mind trust and safety of both partners. Overall, there is growing evidence in the scientific literature that a positive human-animal relationship can bring intrinsic rewards to the animals and thereby benefit animal welfare. Further research is needed on the underlying processes to establish an effective positive human-animal relationship, especially in regard to the type, frequency, and length of human interaction necessary. In particular, the importance of providing animals with a sense of agency over their interactions with humans remains poorly understood.

The dog-human connection

This special issue of The Anatomical Record is the end result of a rare convergence of researchers scattered around the globe who came together to explore the mystery of the dog-human connection. Many of the discussions at the 12th International Congress of Vertebrate Morphology in Prague (July 23, 2019) are echoed within this issue. The enigmatic origins of dog domestication (as well as feralized descendants such as the dingo) are discussed, including phases of domestication that we might infer, and our historical knowledge of dog breeding. Emphasized by the morphological and genetic data are the forces of selection, both unintentional and intentional. In our modern life with dogs, we enjoy their companionship and benefit from the utility of many breeds, but we encounter unintended health care issues that are often breed-specific. Dogs are so different in their sensory specializations (especially olfaction), but have uniquely (among other domestic mammals) developed highly sophisticated means of interspecific communication with humans. In sum, the manuscripts within this issue discuss anatomical, paleontological, genetic, and behavioral evidence bearing on the antiquity of the domestic dog, the process of domestication, and the many ways in which dogs continue to affect human life.

The role of cat eye narrowing movements in cat-human communication

Domestic animals are sensitive to human cues that facilitate inter-specific communication, including cues to emotional state. The eyes are important in signalling emotions, with the act of narrowing the eyes appearing to be associated with positive emotional communication in a range of species. This study examines the communicatory significance of a widely reported cat behaviour that involves eye narrowing, referred to as the slow blink sequence. Slow blink sequences typically involve a series of half-blinks followed by either a prolonged eye narrow or an eye closure. Our first experiment revealed that cat half-blinks and eye narrowing occurred more frequently in response to owners' slow blink stimuli towards their cats (compared to no owner-cat interaction). In a second experiment, this time where an experimenter provided the slow blink stimulus, cats had a higher propensity to approach the experimenter after a slow blink interaction than when they had adopted a neutral expression. Collectively, our results suggest that slow blink sequences may function as a form of positive emotional communication between cats and humans.

Where Do We Stand in the Domestic Dog ( Canis familiaris ) Positive-Emotion Assessment: A State-of-the-Art Review and Future Directions

Although there have been a growing number of studies focusing on dog welfare, the research field concerning dog positive-emotion assessment remains mostly unexplored. This paper aims to provide a state-of-the-art review and summary of the scattered and disperse research on dog positive-emotion assessment. The review notably details the current advancement in dog positive-emotion research, what approaches, measures, methods, and techniques have been implemented so far in emotion perception, processing, and response assessment. Moreover, we propose possible future research directions for short-term emotion as well as longer-term emotional states assessment in dogs. The review ends by identifying and addressing some methodological limitations and by pointing out further methodological research needs.

Methods of Assessment of the Welfare of Shelter Cats: A Review

At any moment, there are millions of cats housed in foster care facilities for abandoned and stray animals for various reasons worldwide. Care, management and regulation among these facilities differ. Moreover, shelters can never substitute the full comfort of a good home for the animal, and the welfare of cats in shelters is a subject of discussion in many respects. Cats are animals sensitive to changes; for most of them, placement in a shelter is a stressful experience because of changes in routine, environment and the presence of other animals. Stress is reflected in changes in behaviour, causes fluctuations in physiological values and disrupts the immune system, which is a predisposition to the development or reactivation of disease. Evaluation of the presence and intensity of negative impacts is possible through the use of evaluation tools based on indicators that help set the environment and management of keeping so as to disrupt the quality of life as little as possible. Although a comprehensive and valid welfare tool that would evaluate animal-based and at the same time resource-based (or management-based) indicators of cats in shelters is not currently available, it is possible to use partial evaluation of individual welfare indicators to assess welfare. This review aims to provide the readers with an insight into current options of assessment of the welfare of cats in shelters with an emphasis on behavioural, physiological and health indicators with an application in both practical and scientific contexts.

Challenges in the comparative study of empathy and related phenomena in animals

The aim of this review is to discuss recent arguments and findings in the comparative study of empathy. Based on a multidisciplinary approach including psychology and ethology, we review the non-human animal literature concerning theoretical frameworks, methodology, and research outcomes. One specific objective is to highlight discrepancies between theory and empirical findings, and to discuss ambiguities present in current data and their interpretation. In particular, we focus on emotional contagion and its experimental investigation, and on consolation and targeted helping as measures for sympathy. Additionally, we address the feasibility of comparing across species with behavioural data alone. One main conclusion of our review is that animal research on empathy still faces the challenge of closing the gap between theoretical concepts and empirical evidence. To advance our knowledge, we propose to focus more on the emotional basis of empathy, rather than on possibly ambiguous behavioural indicators, and we provide suggestions to overcome the limitations of previous research .

Facial expressions of pain in cats: the development and validation of a Feline Grimace Scale

Grimace scales have been used for pain assessment in different species. This study aimed to develop and validate the Feline Grimace Scale (FGS) to detect naturally-occurring acute pain. Thirty-five client-owned and twenty control cats were video-recorded undisturbed in their cages in a prospective, case-control study. Painful cats received analgesic treatment and videos were repeated one hour later. Five action units (AU) were identified: ear position, orbital tightening, muzzle tension, whiskers change and head position. Four observers independently scored (0-2 for each AU) 110 images of control and painful cats. The FGS scores were higher in painful than in control cats; a very strong correlation with another validated instrument for pain assessment in cats was observed (rho = 0.86, p < 0.001) as well as good overall inter-rater reliability [ICC = 0.89 (95% CI: 0.85-0.92)], excellent intra-rater reliability (ICC > 0.91), and excellent internal consistency (Cronbach's alpha = 0.89). The FGS detected response to analgesic treatment (scores after analgesia were lower than before) and a cut-off score was determined (total pain score > 0.39 out of 1.0). The FGS is a valid and reliable tool for acute pain assessment in cats.

Geometric morphometrics for the study of facial expressions in non-human animals, using the domestic cat as an exemplar

Facial expression is a common channel for the communication of emotion. However, in the case of non-human animals, the analytical methods used to quantify facial expressions can be subjective, relying heavily on extrapolation from human-based systems. Here, we demonstrate how geometric morphometrics can be applied in order to overcome these problems. We used this approach to identify and quantify changes in facial shape associated with pain in a non-human animal species. Our method accommodates individual variability, species-specific facial anatomy, and postural effects. Facial images were captured at four different time points during ovariohysterectomy of domestic short haired cats (n = 29), with time points corresponding to varying intensities of pain. Images were annotated using landmarks specifically chosen for their relationship with underlying musculature, and relevance to cat-specific facial action units. Landmark data were subjected to normalisation before Principal Components (PCs) were extracted to identify key sources of facial shape variation, relative to pain intensity. A significant relationship between PC scores and a well-validated composite measure of post-operative pain in cats (UNESP-Botucatu MCPS tool) was evident, demonstrating good convergent validity between our geometric face model, and other metrics of pain detection. This study lays the foundation for the automatic, objective detection of emotional expressions in a range of non-human animal species.

Sex and Age Don't Matter, but Breed Type Does-Factors Influencing Eye Wrinkle Expression in Horses

Identifying valid indicators to assess animals' emotional states is a critical objective of animal welfare science. In horses, eye wrinkles above the eyeball have been shown to be affected by pain and other emotional states. From other species we know that individual characteristics, e.g., age in humans, affect facial wrinkles, but it has not yet been investigated whether eye wrinkle expression in horses is systematically affected by such characteristics. Therefore, the aim of this study was to assess how age, sex, breed type, body condition, and coat colour affect the expression and/or the assessment of eye wrinkles in horses. To this end, we adapted the eye wrinkle assessment scale from Hintze et al. (1) and assessed eye wrinkle expression in pictures taken from the left and the right eye of 181 horses in a presumably neutral situation, using five outcome measures: a qualitative first impression reflecting how worried the horse is perceived by humans, the extent to which the brow is raised, the number of wrinkles, their markedness and the angle between a line through both corners of the eye and the topmost wrinkle. All measures could be assessed highly reliable with respect to intra- and inter-observer agreement. Breed type affected the width of the angle [F (2,114) = 8.20, p < 0.001], with thoroughbreds having the narrowest angle (M = 23.80, SD = 1.60), followed by warmbloods (M = 28.00, SD = 0.60), and coldbloods (M = 31.00, SD = 0.90). None of the other characteristics affected any of the outcome measures, and eye wrinkle expression did not differ between the left and the right eye area (all p-values > 0.05). In conclusion, horses' eye wrinkle expression and its assessment in neutral situations was not systematically affected by the investigated characteristics, except for "breed type", which accounted for some variation in "angle"; how much eye wrinkle expression is affected by emotion or perhaps mood needs further investigation and validation.

Shelter housing for cats: Principles of design for health, welfare and rehoming

Practical relevance: Shelters and rehoming centres are a valuable tool in the population management and rehoming of cats. However, housing large numbers of a relatively asocial species in close proximity poses significant challenges. Well-designed accommodation enables improved standards of husbandry, as well as a better working environment for staff. This can have a significant benefit in expediting rehoming, as cats are healthier, and more likely to display natural behaviors and have positive interactions with potential adopters. Global importance: As cat overpopulation is such a widespread issue, cat shelters are common in many countries. This review will be of interest to those involved in the design and construction of cat shelters, and to those caring for the cats within them. The principles discussed also apply to boarding, breeding, research and hospitalization facilities. Challenges: Shelter housing poses substantial challenges in terms of maintaining positive health and wellbeing while sustaining adoption at an optimum rate. Disease control and biosecurity are typically facilitated by having a relatively barren, easily cleanable environment. However, this must be weighed against the provision of opportunities for cats to perform natural behaviors such as hiding, perching and interacting if they wish. A more enriched environment has also been shown to expedite adoption.

AIMS

This review, the first in a two-part series, discusses practical aspects of housing and shelter design for the health, welfare and adoption of shelter cats. Evidence base: There is a relatively small body of empirical evidence to inform shelter design recommendations. The recommendations in this article are based on a careful review of the available evidence, some of which has come from allied fields such as the care of experimental animals. Where evidence is not yet available, recommendations have been based on field experience and collective expert opinion.

In the Eye of the Beholder: Owner Preferences for Variations in Cats' Appearances with Specific Focus on Skull Morphology

Simple Summary

Recently, there has been an increase in popularity of cats with different skull shapes, including shortened or lengthened muzzles. Skull shape, like other physical features, may affect human preferences; however, it is also more likely to have an impact on the welfare of the cat. We asked people to score their preference for 15 pictures of cats across two surveys. Extreme face shapes (those that were very short or very long) were least preferred. Short-faced cats were less popular amongst cat owners from animal related jobs as opposed to other people. Respondents that had a short or long-faced cat preferred cats with the same skull shape, but also had lower preferences for the opposite skull shape. Respondents from Asia, as compared to those from elsewhere, gave higher preference scores to both long and short-faced cats. Amongst the other features, green eyes, a ginger coat color and medium length coat were most preferred, although the ability to draw conclusions around these features is limited, given they are not necessarily independent of skull shape. This study provides the first evidence that preferences for cat breeds, and their associated skull morphologies, are driven by both culture and owner experience. This information may inform future research concerning the preferences of cat owners.

Abstract

Changes in the popularity of cat breeds are largely driven by human perceptions of, and selection for, phenotypic traits including skull morphology. The popularity of breeds with altered skull shapes appears to be increasing, and owner preferences are an important part of this dynamic. This study sought to establish how and why a range of phenotypic attributes, including skull shape, affect preferences shown by cat owners. Two questionnaires were distributed on-line to cat owners who were asked to rate preferences for pictures of cats on a 0–10 scale. Veterinarian consensus established the skull types of the cats pictured (i.e., level of brachycephaly (BC) or dolichocephaly (DC)). Preferences were then explored relative to cat skull type, coat and eye color, and coat length. Generalized estimating equations identified relationships between physical characteristics and respondent ratings. Further sub-analyses explored effects of respondents’ occupation, location and previous cat ownership on rating scores. Overall, cats with extreme changes in skull morphology (both BC and DC) were significantly less preferred than mesocephalic cats. Green eyes, ginger coat color and medium length coat were most preferred. Current owners of a BC or DC pure bred cat showed significantly greater preference for cats with similar features and significantly lower preference for the opposite extreme. Respondents from Asia were significantly more likely to prefer both BC and DC cats as compared to respondents from other locations. Finally, those in an animal care profession, as compared to other professions, provided a significantly lower preference rating for BC cats but not for DC cats. This work, despite the acknowledged limitations, provides preliminary evidence that preferences for cat breeds, and their associated skull morphologies, are driven by both cultural and experiential parameters. This information may allow for better targeting of educational materials concerning cat breeds.

Dogs and humans respond to emotionally competent stimuli by producing different facial actions

The commonality of facial expressions of emotion has been studied in different species since Darwin, with most of the research focusing on closely related primate species. However, it is unclear to what extent there exists common facial expression in species more phylogenetically distant, but sharing a need for common interspecific emotional understanding. Here we used the objective, anatomically-based tools, FACS and DogFACS (Facial Action Coding Systems), to quantify and compare human and domestic dog facial expressions in response to emotionally-competent stimuli associated with different categories of emotional arousal. We sought to answer two questions: Firstly, do dogs display specific discriminatory facial movements in response to different categories of emotional stimuli? Secondly, do dogs display similar facial movements to humans when reacting in emotionally comparable contexts? We found that dogs displayed distinctive facial actions depending on the category of stimuli. However, dogs produced different facial movements to humans in comparable states of emotional arousal. These results refute the commonality of emotional expression across mammals, since dogs do not display human-like facial expressions. Given the unique interspecific relationship between dogs and humans, two highly social but evolutionarily distant species sharing a common environment, these findings give new insight into the origin of emotion expression.

Facial correlates of emotional behaviour in the domestic cat (Felis catus)

Leyhausen's (1979) work on cat behaviour and facial expressions associated with offensive and defensive behaviour is widely embraced as the standard for interpretation of agonistic behaviour in this species. However, it is a largely anecdotal description that can be easily misunderstood. Recently a facial action coding system has been developed for cats (CatFACS), similar to that used for objectively coding human facial expressions. This study reports on the use of this system to describe the relationship between behaviour and facial expressions of cats in confinement contexts without and with human interaction, in order to generate hypotheses about the relationship between these expressions and underlying emotional state. Video recordings taken of 29 cats resident in a Canadian animal shelter were analysed using 1-0 sampling of 275 4-s video clips. Observations under the two conditions were analysed descriptively using hierarchical cluster analysis for binomial data and indicated that in both situations, about half of the data clustered into three groups. An argument is presented that these largely reflect states based on varying degrees of relaxed engagement, fear and frustration. Facial actions associated with fear included blinking and half-blinking and a left head and gaze bias at lower intensities. Facial actions consistently associated with frustration included hissing, nose-licking, dropping of the jaw, the raising of the upper lip, nose wrinkling, lower lip depression, parting of the lips, mouth stretching, vocalisation and showing of the tongue. Relaxed engagement appeared to be associated with a right gaze and head turn bias. The results also indicate potential qualitative changes associated with differences in intensity in emotional expression following human intervention. The results were also compared to the classic description of "offensive and defensive moods" in cats (Leyhausen, 1979) and previous work by Gourkow et al. (2014a) on behavioural styles in cats in order to assess if these observations had replicable features noted by others. This revealed evidence of convergent validity between the methods However, the use of CatFACS revealed elements relating to vocalisation and response lateralisation, not previously reported in this literature.