Acoustic Vocalisation Early Response Technologies (AVERT)

Laboratory animal welfare is assessed on the basis of visual cues. However, some laboratory species are known to suppress these cues during experiments or when humans are present. One such species is the mouse (Mus musculus), the species most commonly used in regulated procedures across the UK, with almost three million used last year alone. Mice can suffer illness and distress for some time before visible signs become evident, making visual cues (e.g. immobility, grimacing, piloerection) a poor means of welfare assessment. Our reliance on these cues means that laboratory mice are put at risk of prolonged and unnecessary stress, illness and/or pain. For example, in part because they are a social species, mice will not usually be separated from cagemates to prevent conspecific aggression until fight wounds are evident.

Preliminary data from a study by the NC3Rs at MRC Harwell indicated that thousands of animals were wounded by conspecifics in aggressive bouts each year, and that the majority of mice injured had reached humane endpoints. It is therefore necessary to consider alternatives to visual inspection as a means of welfare assessment. Ultrasonic vocalisations (USVs) that indicate pain, distress and aggression are emitted before visible cues become evident. Preliminary testing at MRC Harwell indicates that USVs can be detected using specialised equipment in the home-cage environment. This means that USVs could be assessed without causing additional disturbance relative to standard daily checks.

I intend to deduce links between USVs and particular behaviours indicative of a welfare issue, such as aggression or pain, and build a searchable repository containing this data. This will allow earlier interventions to maximise welfare and could enable the establishment of earlier humane end points. It is hoped that this research could eventually lead to the development of an automated tool, the concept for which will be developed in this project.