Objective graphical clustering of spatiotemporal gait pattern in patients with Parkinsonism

The goal of this study was grouping patients with parkinsonism that share similar gait characteristics based on principal component analysis (PCA). Spatiotemporal gait data during self-selected walking were obtained from 15 patients with Vascular Parkinsonism, 15 patients with Idiopathic Parkinson&#...

Full description

Bibliographic Details
Main Author: Ferreira, Flora José Rocha (author)
Other Authors: Gago, Miguel (author), Mollaei, Nafiseh (author), Bicho, Estela (author), Sousa, Nuno (author), Gama, João (author), Ferreira, Carlos (author)
Format: conferencePaper
Language:eng
Published: 2020
Subjects:
Parkinsonism
Gait analysis
Clustering
Classification
Ciências Médicas::Ciências da Saúde
Science & Technology
Saúde de qualidade
Online Access:http://hdl.handle.net/1822/69786
Country:Portugal
Oai:oai:repositorium.sdum.uminho.pt:1822/69786
Description
Summary:The goal of this study was grouping patients with parkinsonism that share similar gait characteristics based on principal component analysis (PCA). Spatiotemporal gait data during self-selected walking were obtained from 15 patients with Vascular Parkinsonism, 15 patients with Idiopathic Parkinson's Disease and 15 Controls. PCA was used to reduce the dimensionality of 12 gait characteristics for the 45 subjects. Fuzzy C-mean cluster analysis was performed plotting the first two principal components, which accounted for 84.1% of the total variability. Results indicates that it is possible to quantitatively differentiate different gait types in patients with parkinsonism using PCA. Objective graphical classification of gait patterns could assist in clinical evaluation as well as aid treatment planning.

Similar Items