New techniques for pain recognition: What are the applications, where are the limits

CB Johnson^1,2, TJ Gibson^1,2, P Flint², PW Wilson² and DJ Mellor¹

^{1  Animal Welfare Science and Bioethics Centre, College of Sciences, Massey University, Palmerston North, New Zealand
1  Institute of Veterinary, Animal and Biomedical Sciences, College of Sciences, Massey University, Palmerston North, New Zealand}

• Abstract
• Introduction
• Electroencephalographic analysis
• Advantages and disadvantages of electroencephalographic analysis
• Ethological behavioural analysis
• Advantages and disadvantages of behavioural analysis
• Velvet antler removal in red deer
• Conclusions
• References
  
  
• Table 1
• Figure 1
• Figure 2
  
  

Abstract

Electroencephalography and ethological behavioural analysis have proved to be useful in the investigation of pain in animals in a number of applied animal welfare research environments. We have used these techniques in our laboratory to investigate the effects of age on castration pain in lambs, methods of analgesia following velvet antler removal in red deer and surgical dehorning pain in cattle.

This paper will discuss the interpretation of data generated using these two techniques and will consider their limitations together with the typical features of situations where each may be useful. EEG analysis has the advantages that it provides robust data using few animals. Its disadvantages include the need for the noxious stimulus to be acute and of a substantial magnitude. Additionally EEG analysis requires that the animals be anaesthetised. This is an advantage in that it allows pain to be studied without any experimental animals perceiving pain, but a disadvantage in that it makes studies complex and relatively expensive to undertake. Behavioural analysis has the advantages that it is less expensive to undertake and can detect the effects of much more subtle noxious stimuli. Its disadvantages are that it can only be carried out using conscious animals that must perceive the experimentally induced pain and that it requires much larger numbers of animals.

The complementary nature of the information gained from the use of both techniques together will be highlighted using the model of the Natur-O constricting ring for velvet antler removal in young deer (spikers). Other examples of situations where these techniques have uniquely been able to deliver information on pain perception will be mentioned.

Introduction

Of the many research techniques that have utility in investigations of pain, electroencephalographic analysis and ethological behavioural analysis are proving increasingly popular both as individual techniques and more particularly as complementary ways to gain detailed information about animal pain in a variety of applications. Whilst both techniques are complex and have a number of technical requirements and limitations that must be met in order to generate useful data, they both appear very powerful in their ability to deliver specific information about the nature of pain perception in animals. When both techniques are used to investigate the same noxious stimulus, the information can provide unique insights into the nature and degree of that noxious stimulus as well as evaluating potential analgesic strategies.

We have used both of these techniques in a variety of situations in different animal species in order to evaluate the noxious effects of manipulations such as castration, dehorning, velvet antler removal, slaughter and other noxious stimuli. In some of these situations, one of the techniques has proved more useful than the other, but in other situations, particularly castration in lambs and velvet antler removal in red deer, both techniques have been used together to provide unique insights into the perception of the noxious stimuli under investigation (castration in lambs) and different ways of providing analgesia (velvet antler removal in red deer).

This paper will provide a brief description of the methodology, limitations and utility of both techniques, followed by an illustration of how they can be used together. The example of velvet antler removal will be used to illustrate the power of this combination of different techniques in applied animal welfare research.

Electroencephalographic analysis

The electroencephalogram (EEG) is a recording of farfield electrical activity generated by the brain. In mammals, this is considered to represent the activity of the bipolar cells of the cerebral cortex. The EEG is composed of a number of components ranging in frequency from fractions of a Hertz to about two hundred Hertz. Most of the activity occurs between 0.5 and about 30 Hz and this is the range that is usually recorded and analysed. Since the brain is an electrical organ, changes in cerebral function result in changes in the frequency spectrum of the EEG. Such changes may be due to alterations in states of arousal or mentation such as concentration on a stimulus, sleep or anaesthesia. In conscious animals, there are usually too many incidental and artifactual factors which can effect the EEG to allow for useful analysis in analgesia research, but anaesthesia using the minimal anaesthesia model (Murrell and Johnson, 2006), can provide a more controlled state of central nervous system function in which the effects of specific manipulations or noxious stimuli can be quantified. Under these conditions, specific alterations in the EEG can be quantified and reduced to a small number of variables using Fast Fourier Analysis (Cooley and Tukey, 1965) followed by standard statistical techniques that summarise and quantify alterations in frequency spectra. The most commonly reported variables are the median frequency, the spectral edge frequency (usually 95% spectral edge) and the total EEG power. Median frequency (F50) is the statistical median of the power spectrum, 95% spectral edge frequency (F95) is the 95th percentile of the power spectrum and total EEG power (ptot) is the area under the power spectrum curve. For a more detailed description of the derivation of these variables, see Murrell and Johnson (2006) and Johnson (2008).

The three variables, F50, F95 and ptot give different information about changes in the EEG power spectrum. F95 is sensitive to changes at higher frequencies, ptot is sensitive to changes at low frequencies and F50 gives an overall impression of the balance between high and low frequencies. Together these three variables give a good indication of EEG changes. The three variables are derived from short epochs of EEG, usually having durations of one or two seconds. This means that the variables from consecutive EEG epochs can be calculated and plotted to give an indication of changes in the EEG over a period of time. Fig 1 illustrates an example of the way in which changes in these variables can be followed over time taken from a study investigating the EEG effects of surgical dehorning in cattle (Gibson et al., 2007).

Advantages and disadvantages of electroencephalographic analysis

Electroencephalographic analysis requires that animals be anaesthetised using the minimal anaesthesia model (Murrell and Johnson, 2006). This allows the inclusion of a control group without compromising the welfare of the animals involved in the study since all animals can be given analgesia before recovery from anaesthesia. Because of the extreme degree of control required for these studies, significant differences can be seen using relatively small group sizes. Whilst the mathematical concepts involved in the analysis of data are complex, they do allow for the rapid treatment of large amounts of data and provide completely objective analysis. These techniques are more suited to rapidly applied stimuli and give best results with responses in timeframes of seconds to minutes. In addition to this, consistent results have been obtained in an large number of mammalian species making interpretation of new data sets relatively easy. The advantages and disadvantages of electroencephalographic analysis are summarised in Table 1.

Ethological behavioural analysis

An ethogram is the analysis of the behavioural repertoire of an animal over a defined period of time. Within this period, either the proportion of time spent expressing certain behaviours or the number of instances of a behaviour occurring is documented together with the precise times at which these behaviours were expressed. Some behaviours (state behaviours) lend themselves to proportional documentation whilst others (event behaviours) are more suited to numerical recording of individual expressions. For example, postural behaviours such as standing, lying down etc. can best be expressed by proportion whilst movements such as ear flicks can best be expressed as instances. During a given period, a certain animal may lie in sternal recumbency for 20% of the time and flick its ears four times.

The generation of a behavioural ethogram can be done in real time by scoring the animal’s behaviour as it happens, but it is much better to capture the period of behaviour using video and analyse it later. The use of off line video analysis combined with specialist software such as J Watcher (Blumstein and Daniel, 2007) means that a complex ethogram can be created documenting many separate behaviours. In this way a detailed picture of the animal’s behaviour can be built and individual behaviours can be associated together.

For example, it may be found that an animal usually flicks its ears before lying down; such associations can be investigated and documented.
Once a complete ethogram has been created for each animal in a study, the information can be compared for animals in different groups to identify statistical differences between behaviours resulting from different manipulations or treatments.

Advantages and disadvantages of behavioural analysis

Behavioural analysis is a very powerful way of documenting behavioural change in conscious animals. In analgesia research it is limited in that the animals must be conscious throughout the period of study in order to express behaviours and this may place ethical limitations on the inclusion of a control group where a noxious stimulus is applied without analgesia. Because of the inherent variability in behaviour, studies should be carried out on relatively large group sizes and this leads to a lage amount of video footage that can take considerable time to analyse. The data generated by behavioural analysis can give a very accurate indication of changes in an animal’s behaviour over time periods of hours to days, but does not have sufficient temporal resolution to be able to indicate changes in periods as short as seconds to minutes. The advantages and disadvantages of behavioural analysis are summarised in Table 1.

Velvet antler removal in red deer

Velvet antler is often removed from red deer spikers (animals in their first season of antler growth), to allow transport of these animals for slaughter. In New Zealand the Natur-O ring has been provisionally approved as an analgesic technique for this procedure on a temporary basis pending further studies into the efficacy and inherent noxiousness of this technique. On this basis we were asked to investigate both the efficacy and noxiousness of this technique.

Previous studies investigating the noxiousness of velvet antler removal have utilised the minimal anaesthesia model and have clearly demonstrated the noxiousness of antler removal without analgesia and the effectiveness of a lignocaine ring block (Johnson et al 2005). Because of this it was decided to use this approach to compare the use of the Natur-O ring with local anaesthesia. These studies indicated that the Natur-O ring provided analgesia which was comparable to, though not as complete as, that provided by lignocaine. The effect of antler removal following the application of local anaesthetic, Natur-O ring or control (no analgesic technique) is illustrated in Fig 2.

Despite its analgesic efficacy, there remained a question as to the noxiousness of the Natur-O ring following its placement. The Natur-O ring is designed to be placed an hour before antler removal and so there was the possibility that any noxiousness would develop gradually over this period. Because of the relatively long time frame, it was decided to use a behavioural approach to investigate any potential noxiousness. Video was recorded of groups of deer spikers for one hour following the placement of Natur-O ring, local anaesthesia, Natur-O ring and local anaesthesia or no treatment (a control group). Video analysis was performed off line and a detailed ethogram containing 32 separate behaviours constructed. Behaviours include state and event behaviours for the whole animal and also for specific body areas including the head and limbs. These data are still under analysis and are providing an extremely detailed picture of changes in behaviour induced by the application of both Natur-O rings and local anaesthesia.

Conclusions

Electroencephalographic and behavioural analyses are both powerful techniques that can be applied to the investigation of pain in animals. The features and limitations of each technique make them complimentary to each other and mean that together they can be used in a wide variety of circumstances to give a broad and detailed picture of the effects of noxious stimuli and different analgesic techniques designed to mitigate these effects.

References

Blumstein, D.T. and Daniel, J.C. (2007). Quanitifying behaviour the J watcher way. Sinauer Associates, Sunderland MA, USA.

Cooley, J.W. and Tukey, J.W. (1965). An algorithm for the machine calculation of complex Fourier series. Mathematical Computations 19, 297-301.

Gibson, T.J., Johnson, C.B., Stafford, K.J., Mitchinson, S.L, and Mellor, D.J. (2007). Validation of the acute electroencephalographic responses of caves to noxious stimulus with scoop dehorning. New Zealand Veterinary Journal 55 152-157.

Johnson, C.B., Woodbury, W.M., Caulkett, N. and Wilson, P. (2005). Comparison of lidocaine and antler pedicle compression for analgesia during antler removal in red deer (Cervus elaphus) anaesthetised by halothane in oxygen: EEG effects. Journal of Veterinary Anaesthesia and analgesia 32 16-71.

Johnson, C.B. (2008). New approaches to identifying and measuring pain. OIE Technical Series, Vol. 10. Scientific assessment and management of animal pain. Eds D.J. Mellor, P.M. Thornber, A.C.D. Bayvel and S. Kahn (In Press)

Murrell, J.C. and Johnson, C.B. (2006). Neurophysiological techniques to assess pain in animals (Review). Journal of Veterinary Pharmacology and Therapeutics 29, 325-335.

Table 1

Specific features and limitations of EEG and behavioural analysis in the study of pain in animals. 

EEG Analysis	Behavioural Analysis
Animals must be anaesthetised	Animals must be conscious
Statistical differences with small numbers	Larger numbers required for statistical differences
Mathematical concepts complex	No complex mathematical concepts
Rapid analysis of data	Data analysis laborious
Suited to rapidly applied stimuli	Suited to more prolonged perception of pain
Consistent responses in wide variety of mammalian species	Behaviour specific to species and even type
Differentiation of visceral and somatic pain	Behaviour specific to noxious stimuli

Figure. 1

F95, F50 and ptot of the electroencephalogram before and after dehorning in calves dehorned with (grey) or without (black) a lignocaine ring block. Data from Gibson et al. (2007).