2026-10-14T09:32:47Z
https://api.figshare.com/v2/oai
oai:figshare.com:article/33865679
2026-09-16T17:30:58Z
category_4
category_15
category_21
category_122
category_132
portal_5
item_type_3
month_year_09_2026
<p>Classification performances by activity class.</p>
Anne Backes (8614482)
Melanie Eckelt (11741279)
Jennifer Fayad (10887264)
Valeria Serchi (24998600)
Thomas Solignac (24998603)
Tobias Meyer (45322)
Bernd Grimm (11648640)
Caroline Mouton (8838449)
Romain Seil (2742244)
Laurent Malisoux (299532)
Biochemistry
Neuroscience
Biotechnology
Mental Health
Infectious Diseases
wireless instrumented insole
system &# 8217
study therefore aims
normalised confusion matrix
instrumented insoles provide
cohen &# 8217
99 participants performed
using overall accuracy
real &# 8209
inertial sensor data
classify adls within
daily activities using
total activity duration
inertial data
daily living
xlink ">
world monitoring
world environments
unobtrusive solution
strong potential
still limited
step count
semi free
remains difficult
reasonable accuracy
preventive care
outdoor activities
naturalistic conditions
misclassifications occurred
larger cohorts
improved detection
highly accurate
ground truth
fuses pressure
findings demonstrate
excellent performance
direct observation
continuous monitoring
broad set
also highlighting
84 ).
80 ).
10 %).
<div>
<p>Background</p><p>Accurate monitoring of activities of daily living (ADLs) in real‑world environments is essential for preventive care and rehabilitation, yet it remains difficult to achieve outside controlled laboratory settings. Instrumented insoles provide a promising, unobtrusive solution for continuous monitoring. However, the use of multimodal systems integrating plantar pressure and inertial data in naturalistic conditions and larger cohorts is still limited. This study therefore aims to evaluate the accuracy of a wireless instrumented insole (WalkinSense) that fuses pressure and inertial sensor data to classify ADLs within a semi free-living setting.</p>
<p>Methods</p><p>A total of 99 participants performed a broad set of indoor and outdoor activities. Frame-by-frame performance was compared to ground truth (direct observation) using overall accuracy, Cohen’s Kappa, precision, recall, F1-scores and a normalised confusion matrix. Agreement on total activity duration was assessed using mean absolute percentage error (MAPE) scores and Bland-Altman plots.</p>
<p>Results</p><p>Activity classification showed almost perfect agreement (mean overall accuracy 0.87, mean Cohen’s Kappa 0.84). Excellent performance (F1 > 0.90) was achieved for sitting, walking with crutches and cycling, while standing, level and non-level walking showed good performance (F1 > 0.80). Most misclassifications occurred between level walking, hill walking and stairs. Duration-based analysis confirmed high accuracy for sitting, walking with crutches and cycling (MAPE ≤ 10%). Bland-Altman plots indicated overestimation of level walking and underestimation of hill and stair walking. Step count was highly accurate (MAPE < 5%), whereas stair count showed only reasonable accuracy (MAPE ≈ 26%).</p>
<p>Conclusion</p><p>These findings demonstrate the system’s strong potential for real-world monitoring and classification of ADLs while also highlighting the need for improved detection of non-level walking activities.</p></div>
2026-09-16T17:30:46Z
Dataset
Dataset
10.1371/journal.pone.0357961.t001
https://figshare.com/articles/dataset/_p_Classification_performances_by_activity_class_p_/33865679
CC BY 4.0