Enregistré dans:
Détails bibliographiques
Auteurs principaux: Russo, Stefania, Klimas, Rafał, Płonka, Marta, Gall, Hugo Le, Holm, Sven, Stanev, Dimitar, Lipsmeier, Florian, Zanon, Mattia, Kriara, Lito
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2508.05175
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866916885109407744
author Russo, Stefania
Klimas, Rafał
Płonka, Marta
Gall, Hugo Le
Holm, Sven
Stanev, Dimitar
Lipsmeier, Florian
Zanon, Mattia
Kriara, Lito
author_facet Russo, Stefania
Klimas, Rafał
Płonka, Marta
Gall, Hugo Le
Holm, Sven
Stanev, Dimitar
Lipsmeier, Florian
Zanon, Mattia
Kriara, Lito
contents We developed a ResNet-based human activity recognition (HAR) model with minimal overhead to detect gait versus non-gait activities and everyday activities (walking, running, stairs, standing, sitting, lying, sit-to-stand transitions). The model was trained and evaluated using smartphone sensor data from adult healthy controls (HC) and people with multiple sclerosis (PwMS) with Expanded Disability Status Scale (EDSS) scores between 0.0-6.5. Datasets included the GaitLab study (ISRCTN15993728), an internal Roche dataset, and publicly available data sources (training only). Data from 34 HC and 68 PwMS (mean [SD] EDSS: 4.7 [1.5]) were included in the evaluation. The HAR model showed 98.4% and 99.6% accuracy in detecting gait versus non-gait activities in the GaitLab and Roche datasets, respectively, similar to a comparative state-of-the-art ResNet model (99.3% and 99.4%). For everyday activities, the proposed model not only demonstrated higher accuracy than the state-of-the-art model (96.2% vs 91.9%; internal Roche dataset) but also maintained high performance across 9 smartphone wear locations (handbag, shopping bag, crossbody bag, backpack, hoodie pocket, coat/jacket pocket, hand, neck, belt), outperforming the state-of-the-art model by 2.8% - 9.0%. In conclusion, the proposed HAR model accurately detects everyday activities and shows high robustness to various smartphone wear locations, demonstrating its practical applicability.
format Preprint
id arxiv_https___arxiv_org_abs_2508_05175
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Human Activity Recognition from Smartphone Sensor Data for Clinical Trials
Russo, Stefania
Klimas, Rafał
Płonka, Marta
Gall, Hugo Le
Holm, Sven
Stanev, Dimitar
Lipsmeier, Florian
Zanon, Mattia
Kriara, Lito
Machine Learning
68T10 (Primary), 62P10 (Secondary)
I.2.1
We developed a ResNet-based human activity recognition (HAR) model with minimal overhead to detect gait versus non-gait activities and everyday activities (walking, running, stairs, standing, sitting, lying, sit-to-stand transitions). The model was trained and evaluated using smartphone sensor data from adult healthy controls (HC) and people with multiple sclerosis (PwMS) with Expanded Disability Status Scale (EDSS) scores between 0.0-6.5. Datasets included the GaitLab study (ISRCTN15993728), an internal Roche dataset, and publicly available data sources (training only). Data from 34 HC and 68 PwMS (mean [SD] EDSS: 4.7 [1.5]) were included in the evaluation. The HAR model showed 98.4% and 99.6% accuracy in detecting gait versus non-gait activities in the GaitLab and Roche datasets, respectively, similar to a comparative state-of-the-art ResNet model (99.3% and 99.4%). For everyday activities, the proposed model not only demonstrated higher accuracy than the state-of-the-art model (96.2% vs 91.9%; internal Roche dataset) but also maintained high performance across 9 smartphone wear locations (handbag, shopping bag, crossbody bag, backpack, hoodie pocket, coat/jacket pocket, hand, neck, belt), outperforming the state-of-the-art model by 2.8% - 9.0%. In conclusion, the proposed HAR model accurately detects everyday activities and shows high robustness to various smartphone wear locations, demonstrating its practical applicability.
title Human Activity Recognition from Smartphone Sensor Data for Clinical Trials
topic Machine Learning
68T10 (Primary), 62P10 (Secondary)
I.2.1
url https://arxiv.org/abs/2508.05175