Ngueleu AM1, Blanchette AK2, Batcho SC2
1Centre for Interdisciplinary Research in Rehabilitation and Social Integration (CIRRSI), Université Laval, Faculty of Medicine, Quebec, Canada, 2Centre for Interdisciplinary Research in Rehabilitation and Social Integration (CIRRSI), Université Laval, Department of Rehabilitation, Faculty of Medicine, Quebec, Canada
Background: Following daily stride count is crucial in healthy people or people with walking limitations as it enables to prevent sedentary lifestyle and estimate physical activity level, which significantly influences individual's health. Despite the accessibility of several measurement systems, stride count accuracy in real environment remains a major challenge. Instrumented insole seems to present a real potential for an accurate stride count during walking. In the last years, several instrumented insoles were designed and some were used to quantify number of strides. However, validity of most of these insoles were tested on short time (less than 2 minutes) or compared with low number of stride (100 strides). Our previous study seemed to show that one pressure sensor was as accurate as five pressure sensors of insole.
Purpose: Our objective was to evaluate accuracy of instrumented insole integrating one pressure sensor compared to five sensors.
Methods: Based on a scan of subject's foot, five force sensitive resistors (FSR) were placed on key positions of the insole corresponding to five predefined plantar pressure areas. 12 subjects (28.25+3.6 years old, height: 1.70+0.08 cm, weight: 67.83+11.61 kg) wore the instrumented insole and walked during six minutes for each trial at self-selected and maximal speed in indoor and outdoor environments. A team member observed each subject during the walking trials and counted the number of strides taken using a manual stride counter. All trials were also videotaped for direct visual stride count validation as the criterion measured value. Two methods of stride count were tested: count using each FSR and sum of five pressure signals. The validity of our insole was evaluated by Intraclass Correlation Coefficient (ICC) and ANOVA repeated measures. We also used a 2-way ANOVA to test impact of the environments and walking speed on stride count.
Results: For all the subjects, results of stride count were not statistically different using each FSR at self-selected speed (indoor: F=1.93; p= 0.18) and maximal speed (indoor: F=1.03; p= 0.34; outdoor: F=0.55; p= 0.49). Stride counts measured with our insole were high accurate compared to direct visual stride counts for (i) each FSR and (ii) sum of five FSRs at self-selected speed (indoor: (i) 0.95≤ICC ≤0.98, (ii) ICC = 0.94; outdoor: (i) 0.88≤ICC ≤0.99, (ii) ICC = 0.93) and maximal speed (indoor: (i) 0.94≤ICC ≤0.99, (ii) ICC = 0.97; outdoor:
(i) 0.91≤ICC ≤ 0.99
(ii) ICC = 0.95). We observed that the environments (indoor, outdoor) and walking speed affected stride count with p=0.001 and p=0.015 respectively.
Conclusion(s): Stride count with one FSR placed into insole was often more accurate than sum of five FSRs. Considering these accuracies, we thought that stride count with each FSR, individually, could facilitate adaptation and usage of the instrumented insole in different populations.
Implications: The high accuracies of stride count for each FSR placed under different area plantar would be a potential to provide real-time feedback about number of strides taken daily, particularly in stroke survivors presenting diverse walking pattern.
Keywords: Validity, stride count, insole
Funding acknowledgements: Ngueleu was supported by scholarships from the Center for Interdisciplinary Research in Rehabilitation and Social Integration and the consortium MEDTEG-MITACS
Purpose: Our objective was to evaluate accuracy of instrumented insole integrating one pressure sensor compared to five sensors.
Methods: Based on a scan of subject's foot, five force sensitive resistors (FSR) were placed on key positions of the insole corresponding to five predefined plantar pressure areas. 12 subjects (28.25+3.6 years old, height: 1.70+0.08 cm, weight: 67.83+11.61 kg) wore the instrumented insole and walked during six minutes for each trial at self-selected and maximal speed in indoor and outdoor environments. A team member observed each subject during the walking trials and counted the number of strides taken using a manual stride counter. All trials were also videotaped for direct visual stride count validation as the criterion measured value. Two methods of stride count were tested: count using each FSR and sum of five pressure signals. The validity of our insole was evaluated by Intraclass Correlation Coefficient (ICC) and ANOVA repeated measures. We also used a 2-way ANOVA to test impact of the environments and walking speed on stride count.
Results: For all the subjects, results of stride count were not statistically different using each FSR at self-selected speed (indoor: F=1.93; p= 0.18) and maximal speed (indoor: F=1.03; p= 0.34; outdoor: F=0.55; p= 0.49). Stride counts measured with our insole were high accurate compared to direct visual stride counts for (i) each FSR and (ii) sum of five FSRs at self-selected speed (indoor: (i) 0.95≤ICC ≤0.98, (ii) ICC = 0.94; outdoor: (i) 0.88≤ICC ≤0.99, (ii) ICC = 0.93) and maximal speed (indoor: (i) 0.94≤ICC ≤0.99, (ii) ICC = 0.97; outdoor:
(i) 0.91≤ICC ≤ 0.99
(ii) ICC = 0.95). We observed that the environments (indoor, outdoor) and walking speed affected stride count with p=0.001 and p=0.015 respectively.
Conclusion(s): Stride count with one FSR placed into insole was often more accurate than sum of five FSRs. Considering these accuracies, we thought that stride count with each FSR, individually, could facilitate adaptation and usage of the instrumented insole in different populations.
Implications: The high accuracies of stride count for each FSR placed under different area plantar would be a potential to provide real-time feedback about number of strides taken daily, particularly in stroke survivors presenting diverse walking pattern.
Keywords: Validity, stride count, insole
Funding acknowledgements: Ngueleu was supported by scholarships from the Center for Interdisciplinary Research in Rehabilitation and Social Integration and the consortium MEDTEG-MITACS
Topic: Outcome measurement
Ethics approval required: Yes
Institution: Institut de réadaptation en déficience physique de Québec (IRDPQ)
Ethics committee: Research Ethics Committee of CIUSS-CN
Ethics number: CER 2017-577
All authors, affiliations and abstracts have been published as submitted.
