An automated observation system to analyse the behaviour of laboratory rodents

Recording and classifying the behaviour of laboratory rodents is a vital part of a wide range of studies ranging from the discovery of new drugs and detection of harmful side-effects to the biological control of agricultural pests. Until now, it has been a lengthy and painstaking task, requiring human observers to judge, count and record. But a new computerised system developed with EUREKA’s help has done away with the need for human observers, revolutionising the way labs work around the world.

One of the key issues in the biological control of agricultural pests is to find the right natural enemy for a particular pest, for example, parasitic wasps attack pest insects like caterpillars, beetles, aphids, etc. ETHOVISION allows researchers to automate the observation process in the exploratory phase. It measures aspects of the insect’s behaviour in the laboratory (e.g. walking and flight speed, turning rate, time distribution of different areas in space) that have a predictive value for its performance in the field.

A video tracking linked to a computerised image analysis system automatically records and classifies the spatial orientation and movement of an animal, taking away all-too-human frailties such as variation in classification caused by tiredness or different observers.

ETHOVISION systems have now been installed in more than 450 laboratories, including those of many leading pharmaceutical companies, universities and governmental research institutes. In 2001, the product generated sales of around 2.5 million Euro and a 30 per cent increase in turnover for the Dutch company selling it, Noldus Information Technology.

Dr Noldus, managing director of Noldus Information Technology, found his first experience with EUREKA and international cooperation a very rewarding one. “We have been able to take an idea, stemming from the market, through collaborative R&D to a working prototype, and eventually to a very successful commercial product,” he says. “The ’seal of technological excellence’ that comes with the EUREKA status has also been very helpful in our initial marketing efforts. Without the support from EUREKA this probably would not have happened.”

Project partner Utrecht University in the Netherlands provided research in image processing and pattern recognition while the industrial partners (Bayer, Solvay and Lundbeck) played a key role in the high-level functional design, validation and practical testing of the software developed by Noldus in laboratory settings.

As Professor Berry M. Spruijt of Utrecht University says “the project gave our research the advantage of the use of advanced tools for measuring behaviour which we would not have had otherwise. And it has established a long lasting collaboration between the university and Noldus. A win-win situation for all.”