Banca de QUALIFICAÇÃO: VICTOR COSTA DE ANDRADE PIMENTEL

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
STUDENT : VICTOR COSTA DE ANDRADE PIMENTEL
DATE: 25/06/2021
TIME: 15:00
LOCAL: Conferência virtual no Google Meet: meet.google.com/anh-fegb-gmf
TITLE:

Recurrence Quantification Analysis for Gait Disorders Assessment

KEY WORDS:

Human gait, Dynamical gait analysis, Recurrence quantification analysis, Recurrent neural networks, LSTM.

PAGES: 56
BIG AREA: Engenharias
AREA: Engenharia Elétrica
SUMMARY:

The disturbances that may occur in human gait are important biological markers to detect motor disorders caused by neurodegenerative diseases. Although the gait capture techniques are sufficiently accurate nowadays, the gait analysis systems were not yet capable of establishing the correlations between the various factors that influence or alter the normal gait. As gait disorders are related to reduced life expectancy, the current pandemic situation reinforces the urgent need to develop better tools for managing chronic patient conditions. In this sense, recurrence quantification analysis has shown to be a technique capable of gathering useful information about the subtle dynamics of pseudo-periodic systems, such as the human gait. From recurrent quantification measures extracted from gait data it is possible to discriminate between individuals affected by different pathologies related to gait disturbances causes, for example, as so between those who are in different disease stages. Given the above, the execution of an scoping review of the area of nonlinear human gait dynamics analysis for disorders assessment and prediction. The review aims to map the key concepts of the area and clarify its definitions in the literature to identify the types of available evidence in the field, as well as to contribute for the translation of knowledge between the various interdisciplinary areas involving the addressed research field. Moreover, the development of a method for assessment and prediction of gait disorders based on the recurrence quantification measures extracted from time series of locomotor system variables is proposed. To achieve that, it must be developed a LSTM network architecture capable to modelling, from RQA measures, the human gait dynamics for assessment and prediction of disturbances.

 

 

BANKING MEMBERS:
Presidente - 1242315 - PABLO JAVIER ALSINA
Interno - 347628 - ADRIAO DUARTE DORIA NETO
Interno - 2579664 - ALLAN DE MEDEIROS MARTINS
Externa ao Programa - 2179208 - ANA RAQUEL RODRIGUES LINDQUIST
Externo à Instituição - VINICIUS JEFFERSON DIAS VIEIRA - UFPB