On the inverse problem of vibro-acoustography