Do supine oblique views provide better imaging of the cervicothoracic junction than swimmer's views?