A Neural Network for Stance Phase Detection in Smart Cane Users

Caro-Romero, Juan Rafael; Ballesteros, Joaquín; Garcìa-Lagos, Francisco; Urdiales, Cristina; Sandoval, F.

doi:10.1007/978-3-030-20521-8_26

Cited by 4 publications

(5 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A promising solution to these challenges lies in the integration of gait-monitoring sensors into mobility aids commonly used by individuals, such as rollators [13][14][15], and canes [16,17]. This innovative approach ensures that gait analysis is accessible irrespective of location or time, without requiring specific body attachments.…”

Section: Introductionmentioning

confidence: 99%

Examining Gait Characteristics in People with Osteoporosis Utilizing a Non-Wheeled Smart Walker through Spatiotemporal Analysis

Ejaz,

Khan,

Azim

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

Fragility fractures, caused by low-energy trauma, are a significant global health concern, with 158 million people aged 50 and over at risk. Hip fractures, a common issue in elderly patients, are often linked to underlying conditions such as osteoporosis. This study proposed a cost-effective solution using a non-wheeled smart walker with load sensors to measure gait parameters, addressing the high cost of traditional gait analysis equipment, the prototype used PASCO load cells PS2200 for force measurement, eliminating the need for Arduino UNO or microcontroller-based hardware. A lightweight amplifier PS2198 amplified the signal, which was transmitted via USB to a personal computer. PASCO capstone software was used for data recording and visualization. The smart walker was tested on forty volunteers divided into two equal groups: those with osteoporosis and those without, by performing a 10 m walk test three times. ANOVA comparing spatiotemporal parameters (TSPs) of the two participant groups (α = 0.05) showed that significant differences lay in terms of time taken to complete the walk test (p < 0.01), left step length (p = 0.03), walking speed (p = 0.02), and stride length (p < 0.02). The results indicate that this smart walker is a reliable tool for assessing gait patterns in individuals with osteoporosis. The proposed system can be an alternative for time consuming and costly methods such as motion capture, and for socially stigmatizing devices such as exoskeletons. It can also be used further to identify risk factors of osteoporosis.

show abstract

Section: Introductionmentioning

confidence: 99%

Examining Gait Characteristics in People with Osteoporosis Utilizing a Non-Wheeled Smart Walker through Spatiotemporal Analysis

Ejaz,

Khan,

Azim

et al. 2023

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…With the development of machine learning methods, an increasing number of studies have applied machine learning models to applications using inertial sensor data, such as activity recognition [ 23 , 24 ], fall detection [ 25 , 26 ], and gait-phase detection [ 27 , 28 ]. Recently, Caro et al [ 29 ] used a single hidden layer Neural Network (NN) on the data from force sensors on the tip of a cane to estimate the cane contact phase. Because the neural network adapts to multiple gait conditions, it improves the estimation accuracy of the cane contact phase compared to a finite state machine method.…”

Section: System Overviewmentioning

confidence: 99%

Inertial Sensor-Based Instrumented Cane for Real-Time Walking Cane Kinematics Estimation

Fernandez

Ahmad

Wada

2020

Sensors

View full text Add to dashboard Cite

Falls are among the main causes of injuries in elderly individuals. Balance and mobility impairment are major indicators of fall risk in this group. The objective of this research was to develop a fall risk feedback system that operates in real time using an inertial sensor-based instrumented cane. Based on inertial sensor data, the proposed system estimates the kinematics (contact phase and orientation) of the cane. First, the contact phase of the cane was estimated by a convolutional neural network. Next, various algorithms for the cane orientation estimation were compared and validated using an optical motion capture system. The proposed cane contact phase prediction model achieved higher accuracy than the previous models. In the cane orientation estimation, the Madgwick filter yielded the best results overall. Finally, the proposed system was able to estimate both the contact phase and orientation in real time in a single-board computer.

show abstract

“…For ergonomic purposes, the design has paid special attention to the weight distribution so that the positions of the battery and the other electronic components do not affect the center of mass of the cane. The main functions of the Smart Cane are the estimation of the weight-bearing and the detection of the support and non-support periods on the cane [11]. These periods are related to the stance phase on the affected leg [12], and the accuracy of its estimation is directly related to the sensor sampling rate and the method used to estimate them.…”

Section: The Smart Cane Platformmentioning

confidence: 99%

“…Higher sampling rates and complex estimation approaches increase accuracy, but the battery runs out quickly. To estimate the period of support and non-support, two different approaches can be used: a finite state machine with a threshold [7], and a neural network [11]. The simplest approach, the finite state machine, has a higher relative error and lower battery consumption than the neural network because it runs inside of the microcontroller.…”

Section: The Smart Cane Platformmentioning

confidence: 99%

Self-Adaptation of mHealth Devices: The Case of the Smart Cane Platform

Ayala

Ballesteros

Caro-Romero

et al. 2019

13th International Conference on Ubiquitous Computing and Ambient ‪Intelligence UCAmI 2019‬

View full text Add to dashboard Cite

Nowadays, more than one billion people are in need of one or more assistive technologies, and this number is expected to increase beyond two billion by 2050. The majority of assistive technologies are supported by battery-operated devices like smartphones and wearables. This means that battery weight is an important concern in such assistive devices because it may affect negatively its ergonomics. Saving power in these assistive devices is of utmost importance for its potential twofold benefits: extend the device life and reduce the global warming aggravated by billion of these devices. Dynamic Software Product Lines (DSPLs) are a suitable technology that supports system adaptation, in this case, to reduce energy consumption at runtime, considering contextual information and the current state of the device. However, a reduction in battery consumption could negatively affect other quality of service parameters, like response time. Therefore, it is important to trade-off battery saving and these other concerns. This work illustrates how to approach the self-adaptation of smart assistive devices by means of a DSPL-based strategy that optimizes battery consumption taking into account other QoS parameters at the same time. We illustrate our proposal with a real case study: a Smart Cane that is integrated with a DSPL platform, Tanit. Experimentation shows that it is possible to make a trade-off between different quality concerns (energy consumption and relative error). The results of the experiments allow us to conclude that the Tanit approach elongates battery duration of the Smart Cane in one day (an increase of a 6% with a relative error of 1%), so we improve the user quality of experience and reduce the energy footprint with a reasonable relative error.

show abstract

A Neural Network for Stance Phase Detection in Smart Cane Users

Cited by 4 publications

References 20 publications

Examining Gait Characteristics in People with Osteoporosis Utilizing a Non-Wheeled Smart Walker through Spatiotemporal Analysis

Examining Gait Characteristics in People with Osteoporosis Utilizing a Non-Wheeled Smart Walker through Spatiotemporal Analysis

Inertial Sensor-Based Instrumented Cane for Real-Time Walking Cane Kinematics Estimation

Self-Adaptation of mHealth Devices: The Case of the Smart Cane Platform

Contact Info

Product

Resources

About