Swallowing disorders analysis using surface EMG biomarkers and classification models
Roldan-Vasco S, Orozco-Duque A, Orozco Arroyave JR (2023)
Publication Type: Journal article
Publication year: 2023
Journal
Book Volume: 133
Article Number: 103815
DOI: 10.1016/j.dsp.2022.103815
Abstract
The swallowing process involves complex muscle coordination mechanisms, whose alterations are known as dysphagia. Its instrumental diagnosis is performed by invasive and expensive methods. The surface electromyography (sEMG) emerges as an alternative for automated and objective evaluations of dysphagia symptoms. In this paper we consider thirty-one healthy and 29 dysphagic patients who performed swallowing tasks with water, yogurt, saliva and crackers. The sEMG activity was recorded using bilateral channels for masseter, suprahyoid and infrahyoid muscle groups. Two main analyses were performed. Features in time, frequency, time-frequency, and nonlinear dynamics domains were analyzed to find biomarkers suitable to model dysphagia. Additionally, the automatic discrimination of dysphagia was evaluated with three classification scenarios using: (1) individual features, (2) individual muscle groups, and (3) the combination of muscle groups. Time-features domain exhibited a well-defined representation pattern of swallowing, and achieved the highest individual classification performance (AUC>0.8). The two scenarios with muscle groups yielded the best results along the experiments (AUC>0.85). The best classification results are found with the suprahyoid and masseter muscles, in water and saliva intake. As the main result of the study, we proposed a set of sEMG related biomarkers and classification approaches suitable for automatic dysphagia screening, a step forward in the implementation of non-invasive and objective strategies for swallowing evaluation.
Authors with CRIS profile
Involved external institutions
Colombia (CO)How to cite
APA:
Roldan-Vasco, S., Orozco-Duque, A., & Orozco Arroyave, J.R. (2023). Swallowing disorders analysis using surface EMG biomarkers and classification models. Digital Signal Processing, 133. https://doi.org/10.1016/j.dsp.2022.103815
MLA:
Roldan-Vasco, Sebastian, Andres Orozco-Duque, and Juan Rafael Orozco Arroyave. "Swallowing disorders analysis using surface EMG biomarkers and classification models." Digital Signal Processing 133 (2023).
BibTeX: Download