A carregar...
Publicação
Increased Fall Risk Evaluation in the Elderly: A Video-based Approach for Gait Analysis
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 17.34 MB | Adobe PDF |
Autores
Resumo(s)
Worldwide, falls are a major public health problem. Their risk assessment occurs predominantly in
clinical environments, having a great dependency on the healthcare professionals who perform the evalu ation. Hence, the aim of this thesis was to develop a quantitative and objective approach to assess fall risk
through two-dimensional (2D) video gait analysis. Data was acquired from two groups with antagonistic
risk evaluations, a control group and an elderly group with increased fall risk, during two separate activ ities: walking and a standard Time Up and Go (TUG) test. The video sequences acquired were further
pre-processed with the purpose of obtaining human skeletons for each frame of the video, followed by
the computation of gait and time parameters for both activities. The estimated time parameters by the
developed algorithm for the walking activity were compared with those estimated by the gait analysis
Contemplas software, and the mean difference measured among both systems ranged between 0s and
0.02s. In addition, automatically extracted TUG time parameters were validated with those manually
extracted for every TUG test video, and they proved to be strongly correlated (rho > 0.8). The same
parameters were also in agreement with a study from the literature with similar characteristics, in which
the TUG phases duration estimated were within the same range of the study. Furthermore, a Wilcoxon
Rank Sum test was conducted to determine how well a given parameter could distinguish between young
and elderly groups, and results showed a total of 15 out of 22 discriminant parameters (ρ-value < 0.05),
with a greater representation of TUG time parameters (9/15). Apart from video capture, personal in formation such as age, sex, weight, height, history of falls, use of walking aid and presence of motor
diseases were also acquired, and a Pearson correlation was computed between this supplementary data
and the fall risk binary evaluation for each subject. This correlation revealed a strong linear relationship
between age and increased fall risk, corroborated with a coefficient of 0.98. Results demonstrated that
the developed algorithm was able to detect differences between people who had or not risk of falling,
providing promising fall risk indicators that could contribute to a more objective and detailed evaluation
of fall risk.
Descrição
Tese de mestrado integrado, Engenharia Biomédica e Biofísica (Engenharia Clínica e Instrumentação Médica) Universidade de Lisboa, Faculdade de Ciências, 2022
Palavras-chave
avaliação de risco de queda análise de marcha em vídeo esqueletos em 2D idosos teste TUG Teses de mestrado - 2022