polito.it
Politecnico di Torino (logo)

Hardware and Software sEMG-based Analyses for Stimulation Artifact Suppression

Miriam Roccazzella

Hardware and Software sEMG-based Analyses for Stimulation Artifact Suppression.

Rel. Danilo Demarchi, Paolo Motto Ros, Fabio Rossi, Andrea Mongardi. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Elettronica (Electronic Engineering), 2021

[img]
Preview
PDF (Tesi_di_laurea) - Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives.

Download (10MB) | Preview
Abstract:

Functional Electrical Stimulation (FES) is currently one of the most used technique for muscle rehabilitation in order to help the patient in the recovery of voluntary movement by controlling muscle contraction and functional recovery. A physiotherapist-patient system, such as the one used for this thesis, implies the presence of two subjects: the physiotherapist who performs a certain movement and the patient to whom the same movement is induced by electrostimulation. To obtain this result, the stimulation pattern derives from the analysis of the surface ElectroMyoGraphic (sEMG) signal, taken by the physiotherapist during the execution of the movement, which determines the amplitude of the FES impulse. Once extrapolated, the electrostimulation pattern is applied to the second subject (the patient) to induce movement replication. Since the goal is to improve the acquisition front-end leading to a closed-loop application, it is necessary to record the sEMG signal also from the patient during the stimulation in order to be able to analyze the muscle response. To achieve this is necessary to improve the interface between the acquisition device and the human body. As this device is aimed at rehabilitation and constant use during the day, the main features must be wearability, low power consumption being battery-powered, and high sEMG quality during movement execution to properly detect muscle activity and muscle fatigue resulting from electrical stimulation. The first part of the thesis is aimed at optimizing space and power consumption, for this reason the performance of the device with and without the Driven Right Leg (DRL) circuit was investigated evaluating the Signal to Noise Ratio of both configurations. Following different combinations of feedback and output resistors, evaluating the performances in terms of SNR in motion and in rest conditions, it emerged that the contribution of the DRLis useful to stabilize the signal taken in dynamic conditions. The SNR value using the DRL is equal to 23.03 dB, while without its use 19.56 dB were obtained. Therefore, the use of the DRL in the acquisition front-end has been confirmed. The second part of the thesis concerns the detection of muscle fatigue or muscle voluntary response during electrical stimulation. To perform this analysis by sEMG it is necessary to find a way to avoid the presence of stimulation artifacts (SA) that saturate the acquisition channel. For this reason an hardware-based blanking techniques has been carried on. This technique involves switches on the input of the acquisition channel which allow to disconnect the body when the stimulation impulse is detected. Following several tests, it is evident that the SA is removed but in its place new disturbances are introduced due to the change in impedance that occurs when the body is disconnected from the circuit. It was therefore necessary to face the issue by changing approach and moving to software-based SA removal techniques. In particular, it was decided to take the signal from the muscle adjacent to the stimulated one in order to remove the artifact. The choice to analyze the adjacent muscle has been taken since the SA, even though is still present due to the cross-talk between nearby muscles, has a duration about ten times lower. For this reason the software SA removal, based on the subtraction of the SA from the sEMG signal, allows to maintain more information about the muscular response to the electrical stimulation.

Relatori: Danilo Demarchi, Paolo Motto Ros, Fabio Rossi, Andrea Mongardi
Anno accademico: 2021/22
Tipo di pubblicazione: Elettronica
Numero di pagine: 121
Soggetti:
Corso di laurea: Corso di laurea magistrale in Ingegneria Elettronica (Electronic Engineering)
Classe di laurea: Nuovo ordinamento > Laurea magistrale > LM-29 - INGEGNERIA ELETTRONICA
Aziende collaboratrici: NON SPECIFICATO
URI: http://webthesis.biblio.polito.it/id/eprint/21032
Modifica (riservato agli operatori) Modifica (riservato agli operatori)