Szczegóły publikacji
Opis bibliograficzny
Gait analysis in mixed reality for Parkinson’s disease assessment / Daria HEMMERLING, Marta Kaczmarska, Bartłomiej Krawczyk, Miłosz DUDEK, Mateusz DANIOŁ, Paweł JEMIOŁO, Marek WODZIŃSKI, Magdalena Wójcik-Pędziwiatr // Biomedical Signal Processing and Control ; ISSN 1746-8094. — 2025 — vol. 106 art. no. 107659, s. 1-16. — Bibliogr. s. 16, Abstr. — Publikacja dostępna online od: 2025-02-26
Autorzy (8)
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 158666 |
|---|---|
| Data dodania do BaDAP | 2025-04-09 |
| Tekst źródłowy | URL |
| DOI | 10.1016/j.bspc.2025.107659 |
| Rok publikacji | 2025 |
| Typ publikacji | artykuł w czasopiśmie |
| Otwarty dostęp |  |
| Czasopismo/seria | Biomedical Signal Processing and Control |
Abstract
In this research, we investigate gait differences between individuals with Parkinson’s disease (PD) and healthy controls using the Timed Up-and-Go test. We created an application for MR devices, which enables the data acquisition from accelerometers, gyroscopes, and magnetometers allowing us to capture detailed kinematic data during the test, providing valuable insights into motor function and disease progression. The acquired data undergoes transformation into gait parameters, aiding in the differentiation of PD patients from healthy subjects through machine learning algorithms. Additionally, the study assesses the agreement between signals captured by the HL and those from a reference device to ensure measurement accuracy. Statistical analysis is applied to highlight gait parameters that exhibit significant differences between the analyzed groups. The classification metrics in distinguishing between PD and healthy subjects are reported. Best-performing classifier, Random Forest, achieves balanced accuracy = 81.2%, f1 score = 77.8%. This research is a significant advancement in the objective assessment of PD symptoms, mitigating the limitations associated with conventional clinical evaluations.
