Rodents serve as models for the human brain and behavior, and their calls give researchers a window into their mood. Unfortunately, rodent calls are ultrasonic, so a researcher must record a test and then play the audio back at a slower speed to manually categorize and count calls. Software does exist to accomplish this task, but it does not have the ease of use or accuracy that researchers want. Mide, in partnership with Tufts University and Dr. James Reno, propose to automate this task by creating software that will automatically count calls, categorize the calls, and provide confidence measurements to alert the researcher to gradual changes in a subjects performance. The algorithms to identify rodent calls will be based on the WAAVES algorithms developed by Dr. Reno as part of his paper Automating Ultrasonic Vocalization Analyses: The WAAVES Program. The main software interface will be based on Mides existing SlamStick Lab program, which is used in vibration analysis.
