AI Case Study

"Now, a team at the University of Washington that’s investigating the effects of addictive drugs such as alcohol or opioids has developed a piece of software to listen in on the chitchat of rats and mice. Dubbed DeepSqueak, the program transforms audio signals into sonograms, visual representations of the sounds, that can be analysed using AI techniques, similar to those that enable driverless cars ‘see’ their environment. In this context, however, DeepSqueak is allowing researchers to separate the rodents’ calls from background noise.

“The animals have a rich repertoire of calls, around 20 kinds,” said Dr Kevin Coffey, a postdoctoral fellow in the University’s Neumaier lab. Coffey and his colleagues found that male rodents made happy sounds when they were anticipating a reward, such as sugar, or were playing with their male peers. But, when they sensed a female rodent nearby, their vocalisations became more complex, as if they were singing a courtship song. This effect became even more dramatic when they could smell but not see the female, suggesting that they have distinct songs for different stages of courtship.

Thanks to its low cost and convenience, it is hoped that the technique can eventually be used in the investigation of the effects of addictive drugs by monitoring the psychological states of rodents in various stages of withdrawal.

“If scientists can understand better how drugs change brain activity to cause pleasure or unpleasant feelings, we could devise better treatments for addiction,” said laboratory director Prof John Neumaier."