Eye contact in medical examinations using videophones

In Japan, approximately 12,000,000 cellular phones with a videophone function have been distributed and are becoming an important part of the telemedicine infrastructure. By eye movement analysis, we investigated the effectiveness of eye contact using cellular videophones during medical interviews. Three male volunteers acted as first-time patients and were filmed using a high-resolution camera. The video of each of these volunteers was converted into pictures of three different sizes-70 cm x 40 cm (large picture), 26 cm x 20 cm (medium picture), and 14 cm x 12 cm (small picture)-and viewed on a 32-inch video monitor. The large, medium, and small pictures were considered to represent an actual medical interview, an interview via a television conference system, and an interview via a cellular videophone, respectively. Nine doctors watched these pictures, and their eye movements were recorded with a gaze-point recorder. After watching the videos, the doctors evaluated quality of three size pictures. Standard deviation values of the gaze point coordinates decreased with the picture size. However, a very low decrease was observed in the time ratio, during which the gaze point was on the face (eye contact rate); the rate was 92.2% (large picture), 91.8% (medium picture), and 85.1% (small picture). No statistical difference was observed between the values of the medium and small pictures. Subjective evaluations yielded low scores in the small pictures. The relatively high rate of eye contact rate in the small pictures suggests that a cellular videophone will be a good tool for telemedicine, although the usage by which the subjective evaluations can be improved should be determined by doctors.