Objective clinical gait analysis using inertial sensors and six minute walking test



Publication Details

S. Ameli, F. Naghdy, D. Stirling, G. Naghdy & M. Aghmesheh, "Objective clinical gait analysis using inertial sensors and six minute walking test," Pattern Recognition, vol. 63, pp. 246-257, 2017.


The most common side-effect of chemotherapy is fatigue. Because of its impact on Physical Performance Status (PPS), the degree of fatigue is a factor considered in chemotherapy administration. Conventionally, a questionnaire-based method known as the ECOG table, devised by the Eastern Cooperation Oncology Group (ECOG), is employed to assess the chemotherapy-induced fatigue. The approach is qualitative, subjective, inaccurate and prone to error. To achieve a more reliable method, an objective, quantitative and precise method is proposed to assess the PPS of different groups of cancer patients. The approach was developed based on a six-minute walk test (6MWT) during which the kinematic data of 23 body segments were measured using body-mounted inertial sensors. The data streams were subsequently segmented by a clustering algorithm known as ¿minimum-message-length-encoding¿ (MML) producing a Gaussian mixture model (GMM). Several postural states were captured from the model to derive a holistic index representing the PPS of a patient undergoing chemotherapy. The proposed method was validated by applying it to simulated and real data. For the simulation study, a typical gait behaviour simulating post-chemotherapy conditions was devised in consultation with an oncologist while the real data comprised the gait information obtained from 4 cancer patients. The results indicate that the proposed algorithm clearly identifies the characteristics of ambulatory motion affected by chemotherapy and provides a more accurate measure of fatigue that can assist oncologists to make a more objective decision regarding continuation or termination of treatment.

Please refer to publisher version or contact your library.



Link to publisher version (DOI)