[24 Juli 2024] 9th Joint Seminar on Rehabilitation Engineering and Assistive Technology

Pada hari Rabu, 24 Juli 2024 Departemen Teknik Biomedik menyelenggarakan 9^th Joint Seminar on Rehabilitation Engineering and Assistive Technology. Kegiatan ini berlangsung secara online melalui Apk Zoom.

Presentation list:
(1) A Basic Study on Detection of Intention to Change Walking Speed for FES Gait Control
Ryo Kawashima¹, Takashi Watanabe^2
1 Graduate School of Engineering, Tohoku University
² Graduate School of Biomedical Engineering, Tohoku University

(2) Embedded Image Processing of Face Expression for Drowsiness Detection of Safety Driving System
Tazkia Ghifara Aulia, Achmad Arifin, Sukma Firdaus, Norma Hermawan
Biomedical Engineering Department, ITS, Surabaya, Indonesia

(3) A Preliminary Study on Gait Change Detection Based on Estimated Gait Parameters Using Inertial Sensors
Harunori Ando¹, Takashi Watanabe^2
1 Graduate School of Engineering, Tohoku University
² Graduate School of Biomedical Engineering, Tohoku University

(4) A Fuzzy System for Two-Muscle Stimulation Intensity Control of Hybrid FES-Robotic Rehabilitation System: An Experimental Test in Controlling Shoulder Joint Movement
Nabila Shafa Oktavia, Achmad Arifin, Fauza Arrofiqi
Biomedical Engineering Department, ITS, Surabaya, Indonesia

A basic study on detection of intention to change walking speed for FES gait control
Ryo Kawashima¹, Takashi Watanabe^2
1 Graduate School of Engineering, Tohoku University, Japan
² Graduate School of Biomedical Engineering, Tohoku University, Japan

Abstract

Functional Electrical Stimulation (FES) can be useful to restore walking movement of patients with stroke. In FES gait control, it is difficult to control movements reflecting user intentions, such as changing walking speed, ascending or descending stairs. This study focused on detecting the intention to change walking speed in hemiplegics and aimed to estimate the stride speed of non-affected side using machine learning from Inertial Measurement Unit (IMU). The realization of this system would enable the user to walk freely, considering the user’s intentions. In this report, we examined the training method of the stride speed estimation model and its adaptability to speed change gait in healthy walking. Five types of healthy walking (Slow, Moderate, Fast, Speed up, and Speed down) were measured with IMU attached to the foot. Two training methods were tested, using the non-affected foot sensor and Long Short-Term Memory (LSTM) network. The first method was to estimate stride speed at each time until heel-off (HO) on the affected side, using 500 ms of current and past IMU signals as input. The second method was to estimate stride speed at each stride, using the time-normalized IMU signal from the initial contact (IC) on the non-affected side to the HO on the affected side as input. Three types of constant speed walking (Slow, Moderate, and Fast) were used for the training data, and 2 types of walking (Speed up and Speed down) were used for the test data. The results showed that the training method of estimating the stride speed at each time had smaller Mean Absolute Percentage Error (MAPE) for both Speed down and Speed up walking. Improving estimation accuracy is the next step.

Embedded Image Processing of Face Expression for Drowsiness Detection of Safety Driving System
Tazkia Ghifara Aulia, Achmad Arifin, Sukma Firdaus, Norma Hermawan
Biomedical Engineering Department, ITS, Surabaya, Indonesia

Abstract

Traffic accidents in Indonesia mostly occur due to the lack of driver alertness, such as drowsiness. To prevent this, a system that can detect driver drowsiness early is needed. Previous research has shown a correlation between the percentage of eye closure (PERCLOS) and driver drowsiness, utilizing this for detection and prediction. However, driving in real-world conditions is complex due to the varying conditions and habits of drivers, making detection using a single feature prone to misclassification. In this study, a system is designed to detect drowsiness based on facial expressions, focusing on features such as the eyes, eyebrows, and mouth. Data collection and system testing were conducted using a portable driving simulator (PDS). The eyes will be used to calculate the percentage of eye closure (PERCLOS), while the eyebrows will be used to calculate the percentage of eyebrow raising (PEBR). The mouth feature will be used to calculate number of yawns. In this study, deep learning is used to classify the subjects drowsiness level. The resulting drowsiness classification system model achieved an accuracy up to 0.889 with a loss value of 0.5. In practical application, the model can correctly detected dowsiness in subject conditions.

A Preliminary Study on Gait Change Detection Based on Estimated Gait Parameters Using Inertial Sensors
Harunori Ando¹, Takashi Watanabe^2
1 Graduate School of Engineering, Tohoku University, Japan
² Graduate School of Biomedical Engineering, Tohoku University, Japan

Abstract

Detecting gait changes is effective for predicting falls. However, gait change is often accompanied by changes in multiple gait parameters, and there are few studies that evaluate which gait parameters are the important factors causing the gait change. In our previous study, we attempted to estimate the IMU signals associated with body parts that responsible for abnormal gait for the evaluation of gait function, based on the calculation of importance of the index used in the abnormality detection. This study aimed to test the feasibility of estimating the importance of gait parameters affecting gait change detection based on gait parameters estimated from signals of IMU attached to the foot. First, baseline gait data and two types of gait change data were collected: one with only the stride length reduced and the other with only the stride time increased. Using a model based on One Class Support Vector Machine (OCSVM), gait change was detected from the calculated values of stride length and stride time, and which index had a more important influence on the detected change was calculated from permutation importance. The results showed that the experimentally changed gait parameters had higher values of the importance. Then, the analysis of gait data, in which both stride length and stride time changed, showed high importance values for both indices. These results show the feasibility of a method for evaluating the importance of gait indices for gait change. In the future, other gait indices will be added and tested the method.

A Fuzzy System for Two-Muscle Stimulation Intensity Control of Hybrid FES-Robotic Rehabilitation System: An Experimental Test in Controlling Shoulder Joint Movement
Nabila Shafa Oktavia, Achmad Arifin, Fauza Arrofiqi
Biomedical Engineering Department, ITS, Surabaya, Indonesia 

Abstract

Stroke is a disease caused by the blockage or rupture of blood vessels in the brain. One in four people worldwide is at risk of having a stroke, and more than two-thirds of stroke survivors experience motor impairment, particularly in the upper body, which can hinder daily activities. One of the rehabilitation methods for post-stroke patients currently often used is the hybrid exoskeleton-FES method. However, previous studies still use separate stimulation methods for agonist-antagonist muscles, whereas normal agonist-antagonist muscles typically work together with a coactivation method. This study aims not only to design a fuzzy control method for muscle coactivation in hybrid exoskeleton-FES rehabilitation but also to propose a method that can reduce constant parameters from previous studies that yielded varying results across subjects. The exoskeleton will provide physical therapy by exerting external force and be controlled using a proportional controller to follow the trajectory. Meanwhile, functional electrical stimulation (FES) will be controlled using fuzzy logic to provide neuromuscular therapy to both muscles with a coactivation model for shoulder abduction-adduction movements. FES in this system is provided in the form of sub-threshold and supra-threshold stimuli and tested three times each on six subjects within a hybrid system with a sinusoidal path. Analysis is conducted on the RMSE and torque generated in each trial to detect movements and additional forces on the subjects. The test results show that in the hybrid system with sub-threshold stimulation, the RMSE obtained and compared with the RMSE obtained from testing with the exoskeleton without stimulus resulted in values that were not significantly different, indicating that no movement in the shoulder joint was detected as a response to the stimulus. Meanwhile, in the hybrid system with supra-threshold stimulation, a more stable system with a lower RMSE than the testing with sub-threshold stimulus was achieved. Additionally, through the torque graph comparison between the tests with both stimuli, it was found that testing with supra-threshold stimulus was able to reduce oscillations and improve system stability at the determined target angle with the coactivation model.