Yasuhiro TAKEMURA, Jun-ya SATO¹ and Masato NAKAJIMA¹

Keio Leading-edge Laboratory of Science and Technology, #213, 308-10, Ogura, Saiwai-ku, Kawasaki-shi, Kanagawa 212-0054, Japan
¹Faculty of Science and Technology, Keio University, 14-1, Hiyoshi 3-chome, Kouhoku-ku, Yokohama-shi, Kanagawa 223-5822, Japan

(Received August 13, 2004; Accepted November 24, 2004)

A non-restrictive and non-contact respiratory movement monitoring system that finds the boundary between chest and abdomen automatically and detects the vertical movement of each part of the body separately is proposed. The system uses a fiber-grating vision sensor technique and the boundary position detection is carried out by calculating the centers of gravity of upward moving and downward moving sampling points, respectively. In the experiment to evaluate the ability to detect the respiratory movement signals of each part and to discriminate between obstructive and central apneas, detected signals of the two parts and their total clearly showed the peculiarities of obstructive and central apnea. The cross talk between the two categories classified automatically according to several rules that reflect the peculiarities was < 15%. This result is sufficient for discriminating central sleep apnea syndrome from obstructive sleep apnea syndrome and indicates that the system is promising as screening equipment.

Key words: respiratory movement monitoring, sleep apnea syndrome diagnosis, obstructive sleep apnea, central sleep apnea, fiber-grating vision sensor, paradoxical movement, center of gravity, region of interest