An adaptive interval type-2 fuzzy logic framework for classification of gait patterns of anterior cruciate ligament reconstructed subjects

This paper aims to investigate a gait pattern classification system for anterior cruciate ligament reconstructed (ACL-R) subjects based on the interval type-2 fuzzy logic (FL). The proposed system intends to model the uncertainties present in kinematics and electromyography (EMG) data used for gait analysis due to intra- and inter-subject stride-to-stride variability and nature of signals. Four features were selected from kinematics and EMG data recorded through wearable wireless sensors. The parameters for the membership functions of these input features were determined using the data recorded for 12 healthy and ACL-R subjects. The parameters for output membership functions and rules were chosen based on the recommendations from physiotherapists and physiatrists. The system was trained by using steepest descent method and tested for singleton and non-singleton inputs. The overall classification accuracy results show that the interval type-2 FL system outperforms the type-1 FL system in recognizing the gait patterns of healthy and ACL-R subjects.