A new fractal-based kinetic index to characterize gait deficits with application in stroke survivor functional mobility assessment. (July 2019)

• Record Type: Journal Article
• Authors: Tan, Ming-Gui; Ho, Jee-Hou; Goh, Hui-Ting; Ng, Hoon Kiat; Abdul Latif, Lydia; Mazlan, Mazlina
• Is Part Of: Biomedical signal processing and control. Volume 52(2019)
• Page Start: 403
• Page End: 413
• Publication Date: 2019-07
• Subjects: Fractal dimension -- Gait analysis -- Stroke -- sEMG -- TUG -- Classification; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2018.09.014
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: A new kinetic index (K.I.) is proposed to characterize gait deficits. K.I. has strong correlation with the Time Up and Go (TUG) test score. K.I. could classify stroke survivors into different homogeneous subgroups. Implications of the proposed K.I. on clinical assessment are discussed. Abstract: This paper proposes a new Kinetic Index (K.I.) to characterize the gait deficits in stroke survivors. The index is derived from the fractal properties of surface electromyography (sEMG) signals. The objectives of proposing this K.I. are (i) to find the correlation between sEMG fractal properties with TUG test; (ii) to classify stroke survivors into different homogeneous subgroups based on K.I., and (iii) to compare the classification results based on published methods. To achieve these objectives, 30 stroke survivors with different levels of gait impairments were recruited to perform TUG. sEMG signals from Tibialis Anterior (TA) and Gastrocnemius Lateral (GL) were acquired in a 5-meter walk test. Sliding window Higuchi fractal dimension algorithm was applied to sEMG of these TA and GL muscles to determine the fractal properties. Hierarchical cluster analysis was used to classify stroke survivors into different subgroups with (i) conventional multiple category of gait parameters (Approach 1), and (ii) single input by using the proposed K.I. value (Approach 2). Besides that, classification based on stroke survivors TUG score was also applied. Results showed that K.I. has strong correlation with the TUG score. A higher value in K.I. associates with higher TUG score. This suggests K.I. could quantify gait deficits and detect risk of fall in this population. The classification results from the Approach 1 were similar to previous published studies. The gait parameters from Approach 2 showed similar gait patterns to Approach 1. Meanwhile, gait results from classification based on TUG score yielded heterogeneous subgroups. These results suggested that K.I. was able to assess gait severity among stroke survivors and was more efficient (it requires a single input parameter only) to classify stroke survivors into homogeneous subgroups.