Smart Insole: Remote Gait Detection Algorithm Using Pressure Sensors For Toe Walking Rehabilitation

Toe walking is a gait pattern without contacting the heel on the ground. Toe walking is observed as an expansion of a normal developmental phase among toddlers. Still, the continuation of toe walking in older children may cause severe conditions on their foot, ankle, and muscle. Gait rehabilitation can be improved by employing a device that identifies gait patterns, and provides users with biofeedback. In this paper, we introduce Smart Insole platform that remotely monitors walking patterns and provides feedback to users regarding their toe walks. Smart Insole platform detects physical activities using two pressure sensors on the custom-designed insole. In regard to provide a feedback to users, a vibration is triggered after successive toe walks. We evaluated the accuracy of our activity recognition algorithm in a study with 10 healthy subjects. Our proposed algorithm detects toe or normal heel-to-toe walking, sitting, and standing with an average accuracy of 91.3%. Our platform provides orthopedic professionals with accurate and high standard protocols to track activity patterns and outgrow toe walking.