Gravar-mail: Does preference influence performance when reading different sizes of cranial computed tomography?