There are mainly two kinds of diseases in intermittent claudication. One is lumbar spinal canal stenosis (LCS) and the other is arteriosclerosis obliterans (ASO). Whether diagnosing LCS or ASO is critical issue. Wrong diagnosis might cause patient death. Concerning above, this paper presents a novel approach for classification of intermittent claudication from walking motion. The subject with LED makers walks on the treadmill until she or he feels pain. We measure the walking motion by camera and track the LED markers. Treadmill enables to measure walking motion for a long time, and LED maker provides precious position of every joint in the walking. Then, we can get the information such as joint angle trajectory, hemi-foot step, stance and swing phases without any other sensors like foot switch or force plate. We compare walking motions of healthy parson, LCS patients and ASO patients, find their feature and derive criteria for classification of intermittent claudication. From the results, we reveal that average of ankle joint angles and hemi-foot step would be useful for the classification/diagnosis.
