Proportion-based fuzzy gait phase detection using the smart insole

Abstract

Gait phase detection is meaningful in the field of rehabilitation and daily life assistance. In this paper, a novel proportion-based fuzzy (PBF) algorithm is brought forward to achieve better detecting adaptability at different walking speeds and bodyweights, compared with the traditional threshold algorithm. In the PBF algorithm, the plantar pressure measured by each sensor is first divided by the pressure sum of all sensors, transformed into the proportion coefficient. The proportion coefficient represents the relative relationship among the sensors, reflecting the plantar pressure transmission during the walking cycle. Fuzzy logic is employed to smoothly recognize the gait phase and its universe of discourse is consist of the proportion coefficient. A portable smart insole is developed which can independently accomplish the data sampling, processing and wireless transmission. The PBF algorithm and the threshold algorithm are compared experimentally and the results demonstrate the better accuracy and adaptability of the PBF algorithm. Finally, the PBF algorithm is applied to an incomplete paraplegia patient. The abnormities of the patient’s gait are successfully detected, validating the feasibility of the PBF algorithm in the clinical and daily life applications.