Grimace Scales Research Articles

A deep neural network to assess spontaneous pain from mouse facial expressions.

Grimace scales quantify characteristic facial expressions associated with spontaneous pain in rodents and other mammals. However, these scales have not been widely adopted largely because of the time and effort required for highly trained humans to manually score the images. Convoluted neural networks were recently developed that distinguish individual humans and objects in images. Here, we trained one of these networks, the InceptionV3 convolutional neural net, with a large set of human-scored mouse images. Output consists of a binary pain/no-pain assessment and a confidence score. Our automated Mouse Grimace Scale integrates these two outputs and is highly accurate (94%) at assessing the presence of pain in mice across different experimental assays. In addition, we used a novel set of “pain” and “no pain” images to show that automated Mouse Grimace Scale scores are highly correlated with human scores (Pearson’s r = 0.75). Moreover, the automated Mouse Grimace Scale classified a greater proportion of images as “pain” following laparotomy surgery when compared to animals receiving a sham surgery or a post-surgical analgesic. Together, these findings suggest that the automated Mouse Grimace Scale can eliminate the need for tedious human scoring of images and provide an objective and rapid way to quantify spontaneous pain and pain relief in mice.

The composition and initial evaluation of a grimace scale in ferrets after surgical implantation of a telemetry probe.

Reliable recognition of pain is difficult in ferrets as many currently available parameters are non-specific, inconsistent and/or impractical. Grimace scales have successfully been applied to assess pain in different animal species and might also be applicable to ferrets. To compose a Ferret Grimace Scale (FGS), we studied the facial musculature of ferrets and compared lateral photographs of 19 ferret faces at six time points before and after intraperitoneal telemetry probe implantation. We identified the Action Units (AUs) orbital tightening, nose bulging, cheek bulging, ear changes and whisker retraction as potential indicators of pain in ferrets. To evaluate whether these AUs could reliably be used to identify photographs taken before and after surgery, the photographs were scored 0, 1 or 2 (not, moderately or obviously present) by 11 observers that were blinded to the treatment and timing of the photographs. All AU-scores assigned to the photographs taken five hours after surgery were significantly higher compared to their time-matched baseline scores. Further analysis using the weights that were obtained using a Linear Discriminant Analysis revealed that scoring orbital tightening alone was sufficient to make this distinction with high sensitivity, specificity and accuracy. Including weighted scores for nose bulging, cheek bulging and ear change did not change this. As these AUs had more missing values than orbital tightening, their descriptions should be re-evaluated. Including whisker retraction, which had a negative weight, resulted in lower accuracy and should therefore in its current form be left out of the FGS. Overall, the results of this study suggest that the FGS and the AU orbital tightening in particular could be useful in a multifactorial pain assessment protocol for ferrets. However, before applying the FGS in practice, it should be further validated by incorporating more time points before and after applying (different) painful stimuli, and different levels of analgesia.

Local infiltration of the surgical wound with levobupivacaine, ibuprofen, and epinephrine in postoperative pain: An experimental study

The body areas from where sutures are removed later, where wound healing is delayed. Epidural analgesia is the most effective method but could not be used for postoperative pain. Peripheral nerve blockers also provided excellent analgesia but are not effective in postoperative pain. Infiltration of the surgical wound with local anesthetics is decreased postoperative pain by inhibiting transmission of noxious impulses at the site. The objective of the study was to explore the effect of the local infiltration of the surgical wounds with low-dose of levobupivacaine, ibuprofen, and epinephrine over the sutured muscle wound in postoperative pain. Laparotomy was performed in adult rats under isoflurane anesthesia. During surgery, the surgical wounds were infiltrated with 50μL solution containing 0.3% w/v levobupivacaine, 2mg/mL ibuprofen, and 8mg/mL epinephrine (treatment group) and compared to infiltration of that of water for injection (vehicle group) over the sutured muscle wound before skin closing. Postoperative pain was assessed by rodent grimace scales scoring. The study also carried out for measurement for histopathological examinations and the tensile strength of wound. The one-way ANOVA following the Dunnett Multiple comparisons test was used to show significant differences between parameters at 95% level of confidence. The fall in pain started with three-hour post-surgery in the treatment group. At 24h after the end of the successful infiltration, the treatment group had significant reduction of a pain than vehicle group (p=0.048; q=3.527). After three weeks of the wound were closed, a significant improvement of angiogenesis process (p=0.021) and the tensile strength (p=0.019) for the treatment group as compared to baseline. The experimental study was reported that local infiltration of the surgical wound with levobupivacaine, ibuprofen, and epinephrine combination was effective in the postoperative pain and healing of the surgical wounds.

Methods Used to Evaluate Pain Behaviors in Rodents.

Rodents are commonly used to study the pathophysiological mechanisms of pain as studies in humans may be difficult to perform and ethically limited. As pain cannot be directly measured in rodents, many methods that quantify “pain-like” behaviors or nociception have been developed. These behavioral methods can be divided into stimulus-evoked or non-stimulus evoked (spontaneous) nociception, based on whether or not application of an external stimulus is used to elicit a withdrawal response. Stimulus-evoked methods, which include manual and electronic von Frey, Randall-Selitto and the Hargreaves test, were the first to be developed and continue to be in widespread use. However, concerns over the clinical translatability of stimulus-evoked nociception in recent years has led to the development and increasing implementation of non-stimulus evoked methods, such as grimace scales, burrowing, weight bearing and gait analysis. This review article provides an overview, as well as discussion of the advantages and disadvantages of the most commonly used behavioral methods of stimulus-evoked and non-stimulus-evoked nociception used in rodents.

Development of a Piglet Grimace Scale to Evaluate Piglet Pain Using Facial Expressions Following Castration and Tail Docking: A Pilot Study.

Facial expressions are increasingly being used to assess pain in non-human species, including rodents, horses, and lambs. The development of these species-specific grimace scales has allowed for more rapid pain detection, which can lead to better animal welfare if intervention promptly occurs. For grimace scales to ever be used as a stand-alone measure of pain, it is important they correlate with established pain assessment tools, such as behavioral analysis. This preliminary study aimed to determine whether piglets exhibit pain grimacing and if these facial expressions correlate with their behavior. It also assessed and compared the behavior of boar piglets given an analgesic and topical anesthetic prior to surgical castration and tail docking to piglets that did not receive anything for pain relief. Five-day-old male Yorkshire piglets (n = 19) from four pens were randomly assigned, within their pen, to one of five possible treatments: meloxicam (0.4 mg/kg, intramuscularly) + EMLA® cream, meloxicam (0.4 mg/kg, intramuscularly) + non-medicated cream, saline (intramuscularly) + EMLA® cream, saline (intramuscularly) + non-medicated cream, or no treatment prior to surgical castration and tail docking. Piglet behaviors were video recorded for 8 h immediately after castration, as well as for 1 h at 24 h pre- and post-castration. Their individual behaviors were scored continuously for the first 15 min of every hour of video collected. Facial images were also captured across all time points. A Piglet Grimace Scale (PGS) was developed and used by two observers blinded to treatment, time, and procedure to score over 600 piglet faces. All piglets displayed significant behavioral changes up to 7 h post-castration when compared to baseline, and the use of meloxicam and EMLA® cream was not associated with a reduction in painful behaviors. Significantly higher PGS scores were noted at 0, 3, 4, and 5 h post-castration when compared to PGS scores at 7 h and there was no effect of treatment. PGS scores significantly correlated with piglet behavioral activity. The results suggest that the PGS may have utility for pain evaluation in neonatal pigs.

Pain therapy in small pets

Although many advances in pain therapy have been made in recent years, pain therapy is more difficult in the small domestic animal than in cats and dogs. However, there is the ethical obligation that these animals also receive adequate pain therapy. An analgesic is rarely authorized for use in small pets, with pharmacological investigations often lacking and dosages frequently only determined empirically. The small size of the animals often requires a higher dose per kilogram bodyweight compared to cats and dogs. The dosage itself is also difficult to apply in small animals, because many analgesics must be diluted before their use. In addition, frequent manipulation of small animals for analgesic administration induces stress in the patient, which can intensify the pain. In the present article, those analgesics suitable for use in the small domestic animal are described and the indications for the use of the various types of analgesics are explained. A specialized section concentrates on pain detection and algesimetry in the small domestic animal. The detection of pain is much more difficult in small domestic animals. In the last few years so-called "grimace scales" have been developed which are used to assess the facial expression of the animals.

Real-time application of the Rat Grimace Scale as a welfare refinement in laboratory rats.

Rodent grimace scales have been recently validated for pain assessment, allowing evaluation of facial expressions associated with pain. The standard scoring method is retrospective, limiting its application beyond pain research. This study aimed to assess if real-time application of the Rat Grimace Scale (RGS) could reliably and accurately assess pain in rats when compared to the standard method. Thirty-two male and female Sprague-Dawley rats were block randomized into three treatment groups: buprenorphine (0.03 mg/kg, subcutaneously), multimodal analgesia (buprenorphine [0.03 mg/kg] and meloxicam [2 mg/kg], subcutaneously), or saline, followed by intra-plantar carrageenan. Real-time observations (interval and point) were compared to the standard RGS method using concurrent video-recordings. Real-time interval observations reflected the results from the standard RGS method by successfully discriminating between analgesia and saline treatments. Real-time point observations showed poor discrimination between treatments. Real-time observations showed minimal bias (<0.1) and acceptable limits of agreement. These results indicate that applying the RGS in real-time through an interval scoring method is feasible and effective, allowing refinement of laboratory rat welfare through rapid identification of pain and early intervention.

The relationship between the Rat Grimace Scale and mechanical hypersensitivity testing in three experimental pain models.

The assessment of facial expressions associated with pain has been used to evaluate pain in humans and has recently found application in non-human mammals. These so called 'grimace scales' have the potential to be developed into a widely accepted non-invasive method of measuring pain in laboratory rodents. Currently, common methodologies to assess pain rely on nociceptive tests that assess stimulus evoked withdrawal responses. These tests, however, are limited to the assessment of a reflexive response without an affective component. This study aimed to use the recently developed Rat Grimace Scale (RGS) and assess its relationship with a conventional nociceptive test (the application of von Frey filaments). Fifty-two adult, male Wistar rats were randomized to one of five treatment groups: intraplantar carrageenan, intraplantar complete Freund's adjuvant (CFA), plantar incision, anaesthetic control and saline injection control. The RGS and response to mechanical hypersensitivity testing was evaluated at predetermined time points before and after treatment until withdrawal responses returned to baseline levels. The RGS score significantly increased in all pain models. The peak RGS score also coincided with the development of paw hypersensitivity. However, mechanical hypersensitivity persisted after RGS scores returned to baseline. This study confirms that the three pain models induce pain in rodents and showed that peak pain coincided with peak mechanical hypersensitivity. However, mechanical hypersensitivity remained once pain subsided, mimicking the human experience of CFA injection. These findings further our understanding of the roles of, and relationship between, these assays in the assessment of nociception and pain.

Assessment of animal pain and mechanism-based strategies for its reversal

Effective management of pain is critical to the improvement of animal welfare. For this to happen, pain must be recognised and assessed in a variety of contexts. Pain is a complex phenomenon, making reliable, valid, and feasible measurement challenging. The use of facial expressions as a technique to assess pain in non-verbal human patients has been widely utilised for many years. More recently this technique has been developed for use in a number of non-human species: rodents, rabbits, ferrets, cats, sheep, pigs and horses. Facial expression scoring has been demonstrated to provide an effective means of identifying animal pain and in assessing its severity, overcoming some of the limitations of other measures for pain assessment in animals. However, there remain limitations and challenges to the use of facial expression as a welfare assessment tool which must be investigated. This paper reviews current facial expression pain scales (“Grimace Scales”), discussing the general conceptual and methodological issues faced when assessing pain, and highlighting the advantages of using facial expression scales over other pain assessment methods. We provide guidance on how facial expression scales should be developed so as to be valid and reliable, but we also provide guidance on how they should be used in clinical practice.