Detecting Gait Events from Accelerations Using Reservoir Computing
Segmenting the gait cycle into multiple phases using gait event detection (GED) is a well-researched subject with many accurate algorithms. However, the algorithms that are able to perform accurate and robust GED for real-life environments and physical diseases tend to be too complex for their imple...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/19/7180 |
| _version_ | 1797477035191304192 |
|---|---|
| author | Laurent Chiasson-Poirier Hananeh Younesian Katia Turcot Julien Sylvestre |
| author_facet | Laurent Chiasson-Poirier Hananeh Younesian Katia Turcot Julien Sylvestre |
| author_sort | Laurent Chiasson-Poirier |
| collection | DOAJ |
| description | Segmenting the gait cycle into multiple phases using gait event detection (GED) is a well-researched subject with many accurate algorithms. However, the algorithms that are able to perform accurate and robust GED for real-life environments and physical diseases tend to be too complex for their implementation on simple hardware systems limited in computing power and memory, such as those used in wearable devices. This study focuses on a numerical implementation of a reservoir computing (RC) algorithm called the echo state network (ESN) that is based on simple computational steps that are easy to implement on portable hardware systems for real-time detection. RC is a neural network method that is widely used for signal processing applications and uses a fast-training method based on a ridge regression adapted to the large quantity and variety of IMU data needed to use RC in various real-life environment GED. In this study, an ESN was used to perform offline GED with gait data from IMU and ground force sensors retrieved from three databases for a total of 28 healthy adults and 15 walking conditions. Our main finding is that despite its low complexity, ESN is robust for GED, with performance comparable to other state-of-the-art algorithms. Our results show the ESN is robust enough to obtain good detection results in all conditions if the algorithm is trained with variable data that match those conditions. The distribution of the mean absolute errors (MAE) between the detection times from the ESN and the force sensors were between 40 and 120 ms for 6 defined gait events (95th percentile). We compared our ESN with four different state-of-the-art algorithms from the literature. The ESN obtained a MAE not more than 10 ms above three other reference algorithms for normal walking indoor and outdoor conditions and yielded the 2nd lowest MAE and the 2nd highest true positive rate and specificity when applied to outdoor walking and running conditions. Our work opens the door to using the ESN as a GED for applications in wearable sensors for long-term patient monitoring. |
| first_indexed | 2024-03-09T21:12:14Z |
| format | Article |
| id | doaj.art-8e709698e2204330a3de2737b17ccdab |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T21:12:14Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-8e709698e2204330a3de2737b17ccdab2023-11-23T21:44:45ZengMDPI AGSensors1424-82202022-09-012219718010.3390/s22197180Detecting Gait Events from Accelerations Using Reservoir ComputingLaurent Chiasson-Poirier0Hananeh Younesian1Katia Turcot2Julien Sylvestre3Interdisciplinary Institute for Technological Innovation (3IT), Department of Mechanical Engineering, Université de Sherbrooke, Sherbrooke, QC J1K 2R1, CanadaCentre Interdisciplinaire de Recherche en Réadaptation et Intégration Sociale (Cirris), Department of Kinesiology, Université Laval, Quebec, QC G1M 2S8, CanadaCentre Interdisciplinaire de Recherche en Réadaptation et Intégration Sociale (Cirris), Department of Kinesiology, Université Laval, Quebec, QC G1M 2S8, CanadaInterdisciplinary Institute for Technological Innovation (3IT), Department of Mechanical Engineering, Université de Sherbrooke, Sherbrooke, QC J1K 2R1, CanadaSegmenting the gait cycle into multiple phases using gait event detection (GED) is a well-researched subject with many accurate algorithms. However, the algorithms that are able to perform accurate and robust GED for real-life environments and physical diseases tend to be too complex for their implementation on simple hardware systems limited in computing power and memory, such as those used in wearable devices. This study focuses on a numerical implementation of a reservoir computing (RC) algorithm called the echo state network (ESN) that is based on simple computational steps that are easy to implement on portable hardware systems for real-time detection. RC is a neural network method that is widely used for signal processing applications and uses a fast-training method based on a ridge regression adapted to the large quantity and variety of IMU data needed to use RC in various real-life environment GED. In this study, an ESN was used to perform offline GED with gait data from IMU and ground force sensors retrieved from three databases for a total of 28 healthy adults and 15 walking conditions. Our main finding is that despite its low complexity, ESN is robust for GED, with performance comparable to other state-of-the-art algorithms. Our results show the ESN is robust enough to obtain good detection results in all conditions if the algorithm is trained with variable data that match those conditions. The distribution of the mean absolute errors (MAE) between the detection times from the ESN and the force sensors were between 40 and 120 ms for 6 defined gait events (95th percentile). We compared our ESN with four different state-of-the-art algorithms from the literature. The ESN obtained a MAE not more than 10 ms above three other reference algorithms for normal walking indoor and outdoor conditions and yielded the 2nd lowest MAE and the 2nd highest true positive rate and specificity when applied to outdoor walking and running conditions. Our work opens the door to using the ESN as a GED for applications in wearable sensors for long-term patient monitoring.https://www.mdpi.com/1424-8220/22/19/7180gait event detectionreservoir computingecho state networkIMU sensors |
| spellingShingle | Laurent Chiasson-Poirier Hananeh Younesian Katia Turcot Julien Sylvestre Detecting Gait Events from Accelerations Using Reservoir Computing Sensors gait event detection reservoir computing echo state network IMU sensors |
| title | Detecting Gait Events from Accelerations Using Reservoir Computing |
| title_full | Detecting Gait Events from Accelerations Using Reservoir Computing |
| title_fullStr | Detecting Gait Events from Accelerations Using Reservoir Computing |
| title_full_unstemmed | Detecting Gait Events from Accelerations Using Reservoir Computing |
| title_short | Detecting Gait Events from Accelerations Using Reservoir Computing |
| title_sort | detecting gait events from accelerations using reservoir computing |
| topic | gait event detection reservoir computing echo state network IMU sensors |
| url | https://www.mdpi.com/1424-8220/22/19/7180 |
| work_keys_str_mv | AT laurentchiassonpoirier detectinggaiteventsfromaccelerationsusingreservoircomputing AT hananehyounesian detectinggaiteventsfromaccelerationsusingreservoircomputing AT katiaturcot detectinggaiteventsfromaccelerationsusingreservoircomputing AT juliensylvestre detectinggaiteventsfromaccelerationsusingreservoircomputing |