SCIENTIFIC HUB

Araştırmanın izini sürün.

SCIENTIFIC LITERATURE · PPR

Gait Initiation Biomechanics'ten Sabit Durum Yürüyüş Hızını Tahmin Etmek: Sağlıklı ve Parkinson Popülasyonlarında Makine Öğrenme Yaklaşımı

Predicting Steady-State Walking Velocity from Gait Initiation Biomechanics: A Machine-Learning Approach in Healthy and Parkinsonian Populations

Makine çevirisi · Editoryal inceleme bekliyor

· 2026-02-11

DOI: 10.21203/rs.3.rs-8732923/v1

Erişim kaynaktan kontrol edilmeliPreprintKaynağa git ↗

Araştırma özeti

Aşağıdaki metin kaynak makalenin lisansı doğrulanmış orijinal özetidir. cc by

Abstract Background Steady-state walking velocity (SSWV) is a robust indicator of functional mobility, general health, and fall risk. However, its experimental assessment typically requires the analysis of several consecutive walking steps, which exceeds the recording capabilities of the force plate systems commonly used to investigate gait initiation (GI). This methodological limitation restricts the combined evaluation of anticipatory postural control and steady-state gait performance within a single experimental setup. The aim of this study was thus to develop and validate a predictive model capable of estimating SSWV from biomechanical GI parameters measured with a force plate system. Results Thirty-three participants, including young healthy adults, older healthy adults, and patients with Parkinson’s disease, performed a series of GI trials at spontaneous and maximal speeds. Fourteen biomechanical variables, including anticipatory postural adjustments, dynamic stability components, and braking index components, were extracted from force plate recordings and used as inputs for an artificial neural network. To ensure robustness, a bootstrap resampling procedure was applied. The predictive model demonstrated high accuracy across all groups and conditions, with coefficients of determination ranging from 0.83 to 0.99. Bland–Altman analyses revealed strong agreement between predicted and measured SSWV values, with negligible biases and relative limits of agreement generally below 5%. Bayesian analysis further supported the equivalence between the two measurement approaches (BF₀₁=5.29). Conclusions These findings demonstrate that SSWV can be reliably predicted from GI biomechanics alone. This approach enables a comprehensive assessment of postural control and walking performance using a single force plate system, thereby offering substantial methodological advantages for the evaluation of locomotion in both healthy and clinical populations.

Bilimsel değerlendirme

01

Araştırma sorusu

Sorunun kapsamı orijinal çalışma üzerinden değerlendirilir.

02

Yöntem ve örneklem

Çalışma tasarımı ve örneklem için kaynak makaleyi kontrol edin.

03

Temel bulgular

Bu kayıt için editoryal olarak onaylanmış bulgu özeti henüz yayımlanmadı.

04

Sınırlılıklar

Sonuçların genellenebilirliği, yanlılık ve klinik anlamı ayrıca incelenmelidir.

Bibliyografik veriden yöntem, örneklem veya klinik sonuç çıkarımı yapılmaz.