Software para análise e identificação de desvios de marcha em amputados do membro inferior

  • Ana Teresa da Costa Campos (Student)

Student thesis: Master's Thesis


This report describes the work undertaken during the four-month internship at Padrão Ortopédico, a company specialized in the manufacturing and commercialization of prostheses and orthoses, located in Porto, Portugal. The culmination of this project aimed at developing a software capable of collecting gait data from lower limb amputee patients, specifically transfemoral and transtibial amputees, and analyzing various factors such as joint angles and coordinates of specific anatomical landmarks. The ability to walk is crucial in human life. The gait cycle consists of two phases: the stance phase and the swing phase. The joints primarily involved in this process are the hip, knee, and ankle joints. There is a relationship between the angular data of these joints throughout the gait cycle, especially in the sagittal plane. From these graphs, many conclusions can be drawn about the quality of an individual's walking pattern. As locomotion is of utmost importance, there are currently various systems for its evaluation. This can be done visually, by observing the patient and providing immediate feedback, but it can also be performed using objective systems such as image processing techniques. Markers are widely used in motion capture, as they allow the definition of specific anatomical points of interest. In the case of lower limb amputation, gait is significantly compromised due to the replacement of the limb with a prosthesis, which requires substantial adaptation and effort to regain the ability to walk. The most common types of lower limb amputations are transfemoral (above the knee) and transtibial (below the knee). There are several common postural deviations in individuals with these types of amputations, including the vaulting effect, circumduction, lateral trunk flexion, abrupt plantar flexion, knee hyperextension, and excessive hip lift. The software developed to obtain data of the gait of transfemoral and transtibial amputees, along with the web-based analysis application, successfully collected accurate coordinates from seventeen users and identified the presence of up to six gait deviations. The comparison of the data obtained by this system with the Qualisys data at LABIOMEP was crucial to ensure the reliability of the results. Furthermore, a study was conducted to explore the possibility of collecting data without the use of markers, using machine learning techniques and pose estimation, in order to eliminate a step in the procedure.
Date of Award28 Jul 2023
Original languagePortuguese
Awarding Institution
  • Universidade Católica Portuguesa
SupervisorTiago Araújo Safara Silva (Supervisor) & João Paulo Vilas-Boas Soares Campos (Co-Supervisor)


  • Amputation
  • Prosthetics
  • Lower limb
  • Gait
  • Deviations


  • Mestrado em Engenharia Biomédica

Cite this