Visual perception of emotion cues in dogs: a critical review of methodologies.

Comparative studies of human-dog cognition have grown exponentially since the 2000's, but the focus on how dogs look at us (as well as other dogs) as social partners is a more recent phenomenon despite its importance to human-dog interactions. Here, we briefly summarise the current state of research in visual perception of emotion cues in dogs and why this area is important; we then critically review its most commonly used methods, by discussing conceptual and methodological challenges and associated limitations in depth; finally, we suggest some possible solutions and recommend best practice for future research. Typically, most studies in this field have concentrated on facial emotional cues, with full body information rarely considered. There are many challenges in the way studies are conceptually designed (e.g., use of non-naturalistic stimuli) and the way researchers incorporate biases (e.g., anthropomorphism) into experimental designs, which may lead to problematic conclusions. However, technological and scientific advances offer the opportunity to gather much more valid, objective, and systematic data in this rapidly expanding field of study. Solving conceptual and methodological challenges in the field of emotion perception research in dogs will not only be beneficial in improving research in dog-human interactions, but also within the comparative psychology area, in which dogs are an important model species to study evolutionary processes.

Dogs can infer implicit information from human emotional expressions.

The ability to infer emotional states and their wider consequences requires the establishment of relationships between the emotional display and subsequent actions. These abilities, together with the use of emotional information from others in social decision making, are cognitively demanding and require inferential skills that extend beyond the immediate perception of the current behaviour of another individual. They may include predictions of the significance of the emotional states being expressed. These abilities were previously believed to be exclusive to primates. In this study, we presented adult domestic dogs with a social interaction between two unfamiliar people, which could be positive, negative or neutral. After passively witnessing the actors engaging silently with each other and with the environment, dogs were given the opportunity to approach a food resource that varied in accessibility. We found that the available emotional information was more relevant than the motivation of the actors (i.e. giving something or receiving something) in predicting the dogs' responses. Thus, dogs were able to access implicit information from the actors' emotional states and appropriately use the affective information to make context-dependent decisions. The findings demonstrate that a non-human animal can actively acquire information from emotional expressions, infer some form of emotional state and use this functionally to make decisions.

Perception of dynamic facial expressions of emotion between dogs and humans.

Facial expressions are a core component of the emotional response of social mammals. In contrast to Darwin's original proposition, expressive facial cues of emotion appear to have evolved to be species-specific. Faces trigger an automatic perceptual process, and so, inter-specific emotion perception is potentially a challenge; since observers should not try to "read" heterospecific facial expressions in the same way that they do conspecific ones. Using dynamic spontaneous facial expression stimuli, we report the first inter-species eye-tracking study on fully unrestrained participants and without pre-experiment training to maintain attention to stimuli, to compare how two different species living in the same ecological niche, humans and dogs, perceive each other's facial expressions of emotion. Humans and dogs showed different gaze distributions when viewing the same facial expressions of either humans or dogs. Humans modulated their gaze depending on the area of interest (AOI) being examined, emotion, and species observed, but dogs modulated their gaze depending on AOI only. We also analysed if the gaze distribution was random across AOIs in both species: in humans, eye movements were not correlated with the diagnostic facial movements occurring in the emotional expression, and in dogs, there was only a partial relationship. This suggests that the scanning of facial expressions is a relatively automatic process. Thus, to read other species' facial emotions successfully, individuals must overcome these automatic perceptual processes and employ learning strategies to appreciate the inter-species emotional repertoire.

Is a local sample internationally representative? Reproducibility of four cognitive tests in family dogs across testing sites and breeds.

A fundamental precept of the scientific method is reproducibility of methods and results, and there is growing concern over the failure to reproduce significant results. Family dogs have become a favoured species in comparative cognition research, but they may be subject to cognitive differences arising from genetic (breeding lines) or cultural differences (e.g. preferred training methods). Such variation is of concern as it affects the validity and generalisability of experimental results. Despite its importance, this problem has not been specifically addressed to date. Therefore, we aimed to test the influence of three factors on reproducibility: testing site (proximal environment), breed and sex (phenotype). The same experimenter tested cognitive performance by more than 200 dogs in four experiments. Additionally, dogs' performance was tested in an obedience task administered by the owner. Breed of dog and testing site were found to influence the level of performance only mildly, and only in a means-end experiment and the obedience task. Our findings demonstrate that by applying the same test protocols on sufficiently large samples, the reported phenomena in these cognitive tests can be reproduced, but slight differences in performance levels can occur between different samples. Accordingly, we recommend the utilisation of well-described protocols supported by video examples of the whole experimental procedure. Findings should focus on the main outcome variables of the experiments, rather than speculating about the general importance of small or secondary performance outcomes which are more susceptible to random or local noise.

A survey of the impact of owning a service dog on quality of life for individuals with physical and hearing disability: a pilot study.

BACKGROUND: Quality of life refers to a person's experienced standard of health, comfort and happiness and is typically measured using subjective self-report scales. Despite increasing scientific interest in the value of dogs to human health and the growing demand for trained service dogs, to date no research has reported how service dogs may affect client perceptions of quality of life. METHOD: We compared quality of life scores on the 16 item Flanagan quality of life scale from individuals who owned a trained service dog with those who were eligible to receive a dog, but did not yet have one (waiting list control). Data were analysed separately from two groups; those with a service dog trained for individuals with physical disabilities (with physical service dog: n = 72; waiting for a service dog: n = 24; recruited from Dogs for Good database) and those with a hearing service dog (with hearing service dog = 111; waiting for a service dog = 30; recruited from Hearing Dogs for Deaf People database). RESULTS: When controlling for age and gender individuals scored higher on total quality of life scores if they owned a service dog or a hearing service dog, but this was only statistically significant for those with a service dog. Both groups (physical service dog and hearing service dog) scored significantly higher on items relating to health, working, learning and independence if they owned a service dog, in comparison to those on the waiting list. Those with a physical service dog also scored significantly higher on items relating to recreational activities (including items relating to reading/listening to music, socialising, creative expression), and those involving social interactions (including items relating to participating in organisations, socialising, relationship with relatives). Additionally, those with a physical service dog scored higher on understanding yourself and material comforts than those on the waiting list control. In contrast, those with a hearing service dog appeared to receive fewer benefits on items relating to social activities. CONCLUSIONS: Owning a service dog can bring significant specific and potentially general benefits to the quality of life of individuals with physical disabilities and hearing impairments. These benefits may have considerable implications for individuals with disabilities, society and the economy by promoting independence, learning and working abilities.

The use of imepitoin (Pexion™) on fear and anxiety related problems in dogs - a case series.

BACKGROUND: Fear and anxiety based problems are common in dogs. Alongside behaviour modification programmes, a range of psychopharmacological agents may be recommended to treat such problems, but few are licensed for use in dogs and the onset of action of some can be delayed. The low affinity partial benzodiazepine receptor agonist imepitoin (Pexion™, Boehringer Ingelheim) is licensed for treating canine epilepsy, has a fast onset of action in dogs and has demonstrated anxiolytic properties in rodent models. This case series reports on the use of imepitoin in a group of dogs identified as having fear/anxiety based problems. Twenty dogs were enrolled into the study, attended a behaviour consultation and underwent routine laboratory evaluation. Nineteen dogs proceeded to be treated with imepitoin orally twice daily (starting dose approximately 10 mg/kg, with alterations as required to a maximum 30 mg/kg) alongside a patient-specific behaviour modification plan for a period of 11-19 weeks. Progress was monitored via owner report through daily diary entries and telephone follow-up every two weeks. A Positive and Negative Activation Scale (PANAS) of temperament was also completed by owners during baseline and at the end of the study. RESULTS: The primary outcome measure was average weekly global scores (AWG) from the owner diaries. Average weekly reaction scores (AWR) for each type of eliciting context was used as a secondary outcome. Seventeen dogs completed the trial. Treatment with imepitoin alongside a behaviour modification programme resulted in owner reported improvement with reduced AWG and reduced AWR for anxiety across a range of social and non-social eliciting contexts including noise sensitivities. Significant improvement was apparent within the first week of treatment, and further improvements seen at the 11 week review point. There was a significant reduction in negative activation (PANAS) with 76.5% of owners opting to continue imepitoin at their own expense after completion of the study. CONCLUSIONS: This study provides initial evidence indicating the potential value of imepitoin (Pexion™) alongside appropriate behaviour modification for the rapid alleviation of signs of fear and anxiety in dogs. Further research with a larger subject population and a placebo control would be useful to confirm the apparent efficacy reported here.

On the origin of mongrels: evolutionary history of free-breeding dogs in Eurasia.

Although a large part of the global domestic dog population is free-ranging and free-breeding, knowledge of genetic diversity in these free-breeding dogs (FBDs) and their ancestry relations to pure-breed dogs is limited, and the indigenous status of FBDs in Asia is still uncertain. We analyse genome-wide SNP variability of FBDs across Eurasia, and show that they display weak genetic structure and are genetically distinct from pure-breed dogs rather than constituting an admixture of breeds. Our results suggest that modern European breeds originated locally from European FBDs. East Asian and Arctic breeds show closest affinity to East Asian FBDs, and they both represent the earliest branching lineages in the phylogeny of extant Eurasian dogs. Our biogeographic reconstruction of ancestral distributions indicates a gradual westward expansion of East Asian indigenous dogs to the Middle East and Europe through Central and West Asia, providing evidence for a major expansion that shaped the patterns of genetic differentiation in modern dogs. This expansion was probably secondary and could have led to the replacement of earlier resident populations in Western Eurasia. This could explain why earlier studies based on modern DNA suggest East Asia as the region of dog origin, while ancient DNA and archaeological data point to Western Eurasia.

Dogs with separation-related problems show a "less pessimistic" cognitive bias during treatment with fluoxetine (Reconcile™) and a behaviour modification plan.

BACKGROUND: Canine separation-related problems (SRP) (also described as "separation anxiety" or "separation distress") are among the most common behavioural complaints of dog owners. Treatment with psychoactive medication in parallel with a behaviour modification plan is well documented in the literature, but it is unknown if this is associated with an improvement in underlying affective state (emotion and mood) or simply an inhibition of the behaviour. Cognitive judgement bias tasks have been proposed as a method for assessing underlying affective state and so we used this approach to identify if any change in clinical signs during treatment was associated with a consistent change in cognitive bias (affective state). Five dogs showing signs of SRP (vocalising - e.g. barking, howling-, destruction of property, and toileting - urination or defecation- when alone) were treated with fluoxetine chewable tablets (Reconcile™) and set on a standard behaviour modification plan for two months. Questionnaires and interviews of the owners were used to monitor the clinical progress of the dogs. Subjects were also evaluated using a spatial cognitive bias test to infer changes in underlying affect prior to, and during, treatment. Concurrently, seven other dogs without signs of SRP were tested in the same way to act as controls. Furthermore, possible correlations between cognitive bias and clinical measures were also assessed for dogs with SRP. RESULTS: Prior to treatment, the dogs with SRP responded to ambiguous positions in the cognitive bias test negatively (i.e. with slower running speeds) compared to control dogs (p < 0.05). On weeks 2 and 6 of treatment, SRP dogs displayed similar responses in the cognitive bias test to control dogs, consistent with the possible normalization of affect during treatment, with this effect more pronounced at week 6 (p > 0.05). Questionnaire based clinical measures were significantly correlated among themselves and with performance in the cognitive bias test. CONCLUSION: These results demonstrate for the first time that the clinical treatment of a negative affective state and associated behaviours in a non-human species can produce a shift in cognitive bias. These findings demonstrate how the outcome of an intervention on a clinical problem can be evaluated to determine not only that the subject's behaviour has improved, but also its psychological state (welfare).

Using sniffing behavior to differentiate true negative from false negative responses in trained scent-detection dogs.

False negatives are recorded in every chemical detection system, but when animals are used as a scent detector, some false negatives can arise as a result of a failure in the link between detection and the trained alert response, or a failure of the handler to identify the positive alert. A false negative response can be critical in certain scenarios, such as searching for a live person or detecting explosives. In this study, we investigated whether the nature of sniffing behavior in trained detection dogs during a controlled scent-detection task differs in response to true positives, true negatives, false positives, and false negatives. A total of 200 videos of 10 working detection dogs were pseudorandomly selected and analyzed frame by frame to quantify sniffing duration and the number of sniffing episodes recorded in a Go/No-Go single scent-detection task using an eight-choice test apparatus. We found that the sniffing duration of true negatives is significantly shorter than false negatives, true positives, and false positives. Furthermore, dogs only ever performed one sniffing episode towards true negatives, but two sniffing episodes commonly occurred in the other situations. These results demonstrate how the nature of sniffing can be used to more effectively assess odor detection by dogs used as biological detection devices.

Impulsive for life? The nature of long-term impulsivity in domestic dogs.

Individual differences in impulsivity occur at a cognitive and/or behavioural level and are associated with differing life outcomes. However, there is a lack of empirical evidence to support the long-term stability of these characteristics in non-human animals. This study reports on the stability of convergent measures of impulsivity in domestic dogs assessed more than 6 years apart. Measures were (1) owner assessment by means of a questionnaire, the validated 'Dog Impulsivity Assessment Scale' (DIAS) and (2) dogs' performance in a delayed reward choice test. Dogs had 15-min free access to two food dispensers, one dispensing a piece of food immediately, the other dispensing three pieces after a delay, which increased by 1 s every other time the dogs sampled it. Maximum delay reached in this task reflects decision-making, or cognitive impulsivity, whereas the rate of extra presses on the delayed reward device during the delay can be considered as a measure of motor or behavioural impulsivity. DIAS scores were strongly and significantly correlated across years. The maximum delay reached in the behaviour test was also highly stable, whereas paw-pressing rate was uncorrelated between the years. These results demonstrate that cognitive but not motor impulsivity is highly consistent over time in dogs.

"Have You Seen This Drivel?" A Comparison of How Common Health Issues Are Discussed within Brachycephalic and Non-Brachycephalic Dog Breed Groups on Social Media.

As brachycephalic dog breed ownership increases, there is a growing concern for the welfare of these breeds due to extreme brachycephalism. Understanding the motivations and behaviours of those choosing to own these breeds is important if we wish to address these concerns. The aim of this study was to investigate how owners of brachycephalic and non-brachycephalic dog breeds use social media dog breed groups to discuss common health issues. The purpose of Facebook posts in relation to common health issues, owner awareness of health issues and the role of Facebook facilitated social support were explored between brachycephalic and non-brachycephalic dog owners. Twelve Facebook breed goups were selected (brachycephalic breed groups, n = 6, non-brachycephalic breed groups, n = 6). Using key word searches we extracted the first twenty posts from each group. Thematic analysis revealed three overarching themes: advice seeking, advice giving and community bonding mechanisms. Whilst the purpose of posting did not differ between groups, non-brachycephalic owners appeared to display greater awareness of breed-specific health issues, whilst social support played a more prominent role in brachycephalic groups. This research highlights that social media groups can act as platforms for knowledge exchange and emotional support. These could be utilised by owners, veterinarians and welfare organisations to more effectively enhance dog health and wellbeing.

Working dogs in dynamic on-duty environments: The impact of dark adaptation, strobe lighting and acoustic distraction on task performance.

Sudden changes in sound and light (e.g., sirens and flashing police beacons) are a common component of working dogs' on-duty environment. Yet, how such stimuli impact dogs' ability to perform physical and cognitive tasks has not been explored. To address this shortcoming, we compared the accuracy and time taken for twelve dogs to complete a complex physical and cognitive task, before, during and after exposure to three 'real-world' stimuli: an acoustic distractor (85dB), white strobe lighting (5, 10 & 15 Hz), and exposure to a dazzling white, red, or blue lights. We found that strobe lighting, and to a greater extent, acoustic distraction, significantly reduced dogs' physical performance. Acoustic distraction also tended to impair dogs' cognitive performance. Dazzling lights had no effect on task performance. Most (nine out of twelve) dogs sensitised to the acoustic distraction to the extent of non-participation in the rewarded task. Our results suggest that without effective distractor response training, sudden changes in noise and flickering lights are likely to impede cognitive and physical task performance in working dogs. Repeated uncontrolled exposure may also amplify these effects.

Links between pet ownership and exercise on the mental health of veterinary professionals.

Background: Connections between the effects of pet ownership and exercise on mental health have previously been demonstrated in different populations. However, little is known about the potential effects of pet ownership and exercise on the mental health of veterinary professionals. Since these individuals have a high prevalence of poor mental health and suicide, while they deal with pets professionally, we investigated the impact of pet ownership, exercise and different types of pet ownership on this demographic group. Method: Veterinary professionals over 18 years old answered an online questionnaire about pet ownership, exercise, mental health (including anxiety, depression and suicidal ideation) and mental health correlates. Regression models were used to identify variables significantly related to mental health outcomes. Results: Of 1087 respondents, pet owners were more depressed than non-owners, while anxiety or suicidal ideation was not associated with pet ownership. Dog and horse owners were psychologically healthier (less anxiety, less suicidal ideation) than non-owners of these species. Veterinary professionals who ran regularly had lower anxiety and depression. Those who walked regularly and spent less time sitting experienced fewer depression symptoms. Conclusions: Running, walking and avoiding prolonged sitting might protect the mental health of veterinary professionals. The type of pet owned may be an important factor in the relationship between pet ownership and mental health; however, generally, pet ownership was associated with worse mental health outcomes in this demographic group. Future studies should determine the causal nature of these relationships.

Owners' Beliefs regarding the Emotional Capabilities of Their Dogs and Cats.

The correct interpretation of an animal's emotional state is crucial for successful human-animal interaction. When studying dog and cat emotional expressions, a key source of information is the pet owner, given the extensive interactions they have had with their pets. In this online survey we asked 438 owners whether their dogs and/or cats could express 22 different primary and secondary emotions, and to indicate the behavioral cues they relied upon to identify those expressed emotions. Overall, more emotions were reported in dogs compared to cats, both from owners that owned just one species and those that owned both. Although owners reported a comparable set of sources of behavioral cues (e.g., body posture, facial expression, and head posture) for dogs and cats in expressing the same emotion, distinct combinations tended to be associated with specific emotions in both cats and dogs. Furthermore, the number of emotions reported by dog owners was positively correlated with their personal experience with dogs but negatively correlated with their professional experience. The number of emotions reported in cats was higher in cat-only households compared to those that also owned dogs. These results provide a fertile ground for further empirical investigation of the emotional expressions of dogs and cats, aimed at validating specific emotions in these species.

Trait-Level Resilience in Pet Dogs-Development of the Lincoln Canine Adaptability Resilience Scale (L-CARS).

The concept of psychological resilience is well-explored in the human literature and is often described as the ability to 'bounce back' following adversity. However, it remains a neglected research area in dogs despite observations that like humans, dogs vary in their ability to cope with stress. This study aimed to develop the first canine 'resilience' scale. An on-line survey was developed for owners. This covered demographics, medical/behavioural history of the dog, and 19 potential resilience items assessed using a 5-point Likert scale; 1084 complete responses were received during the survey period, with 329 respondents subsequently completing the questionnaire a second time, 6-8 weeks later. Intra-rater reliability was assessed, and only reliable items retained. A principal component analysis (PCA) with varimax rotation was then performed with components extracted on the basis of the inspection of scree plots and the Kaiser criterion. Items were retained if they loaded >0.4 onto one of the components but removed if they cross-loaded onto more than one component. This resulted in a 14-item, 2-component solution. One component appeared to describe "Adaptability/behavioural flexibility" and the other "Perseverance", which are described in the human literature on resilience. Predictive validity was established for expected correlates, such as problem behaviour. The resulting instrument was called the Lincoln Canine Adaptability and Resilience Scale (L-CARS) and is the first to be developed for the assessment of resilience in dogs.

An Assessment of Scientific Evidence Relating to the Effect of Early Experience on the Risk of Human-Directed Aggression by Adult Dogs.

Human-directed aggression by domestic dogs is a major worldwide public health problem. The causes of aggression are complex, and research in this area often has to balance ecological validity with pragmatic controls; accordingly, it often does not meet the thresholds for quality typically used in reviews applying a classical "evidence-based" approach. Here, we propose a method of literature assessment that makes the "best use" of available evidence to identify and synthesise evidence relating to the most likely risk factors reported in the scientific literature. We used a systematic review process to initially identify relevant literature relating to potential early life experience (i.e., in the first six months of life) risk factors in the dog for human-directed aggression in the adult animal. Fourteen papers met our initial screening process and were subsequently analysed in detail, with data extracted and effect sizes calculated where possible. This highlighted the potential importance of the source of the animal, the age at which it was rehomed, the reason for the acquisition, the experience level of the owner, the animal's socialisation experiences, the consistent husbandry and management practices, the training, the sex ratio of the litter and the history of dogs that display aggression in the pedigree as risk factors. Taken together, it seems that early experiences which limit the ability to develop effective coping strategies and routines may be particularly important. We provide guidance for the future standardised reporting of risk related to human-directed aggression by dogs to allow greater synthesis of the literature in the future.

How Do Dogs Behave When Presented with Situations of Different Emotional Valences?

Dogs are good models for studying behaviour and cognition as they have complex social capabilities. In the current study, we observed how human emotional valences (positive, neutral and negative) affected aspects of dogs' behaviour. We expected that dogs would exhibit more approaching behaviours in the positive condition and more signs of avoidance in the negative one. We analysed videos of 70 adult pet dogs of various breeds taken from an experiment in which one of two actors expressed an emotion and dogs could freely explore the environment for 30 s. Our results show that dogs exhibit differential behaviour when presented with different emotional valences. Two behaviours arose that might be linked to a reciprocal positive emotional state in dogs: tail raised between 90° and 180° and physical contact during sniffing. These behaviours are associated with an active search for information. In the positive conditions, dogs were more willing to explore the social environment and gather information from the actors.

Changes in Dog Behaviour Associated with the COVID-19 Lockdown, Pre-Existing Separation-Related Problems and Alterations in Owner Behaviour.

During the COVID-19 pandemic, lockdowns provided an opportunity to assess what factors, including changes in an owner's routine and time spent at home, were associated with changes in dog behaviour. We undertook a longitudinal survey over a period of 8 months during which we asked about people's work patterns, dog management, and their dogs' behaviour. Generalized linear models revealed that the pre-existence of signs of potential separation-related problems, and especially vocalisation, self-injury, and chewing to escape confinement, was associated with an increase in a range of separation issues. Dogs showing separation-related signs prior to COVID were also more likely to develop more problems during lockdown. Management changes tended to result in increased physical and social stress, with a range of potential compensatory actions taken by the dog, however these signs of stress did not generally appear to be connected to separation-related issues. Survival analysis was used to investigate the emergence of specific issues over time. This indicated that a change to working from home was related initially to a decreased risk of aggression towards the owner, but over time, those who continued to work from the home were at an increased risk of this problem. No other significant time-related relationships were found.

Dog owner mental health is associated with dog behavioural problems, dog care and dog-facilitated social interaction: a prospective cohort study.

Despite numerous qualitative and cross-sectional studies investigating how dog-related factors may impact owners' well-being, empirical studies to test these causal effects are lacking. This prospective cohort study examined the correlation and potential causal effect of 17 dog-related factors with six well-being outcomes (depression, anxiety, loneliness, suicidal ideation, hedonic well-being and eudaimonic well-being) in dog owners. Over a four-week period, 709 adult dog owners reported their weekly well-being and occurrence of each dog-related factor (e.g. how many times they ran with their dogs). A random intercept cross-lagged panel model (RI-CLPM) with significance threshold set at 0.001 was used. Six factors correlated with poorer owner well-being (i.e. aggressive dog behaviour, fearful dog behaviour, poor dog health, failure to provide for the dog, lack of control over the dog, and dog presence). Only 'friendly conversation with others due to the dog' correlated with better well-being. Purposeful reductions in the frequency of dog behavioural and health-related issues are likely to improve owner well-being, as well as greater consistency in dog care (i.e. provide for the dog) and more engagement in friendly dog-facilitated social interactions. No potential causal effects were significant. Further studies investigating causal relationships are essential to improve people's well-being through dog ownership.

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.