Impairment of lower limbs is one of the most evident features in Parkinson's Disease (PD) that represents a heavy burden on the quality of life of people affected by this chronic and neurodegenerative disease. In fact, in PD there is a progressive and general degeneration of the motor control, in which serious problems involve lower limbs, for example during gait, characterized by episodes of freezing of the gait (FOG), bradykinesia or festination; but also disorders of balance, asymmetric postural attitude and postural instability. All these typical manifestations contribute to increase the risk of falls, consequently hindering the execution of the simplest activities in daily living. This technical report is related to Leg Agility, one of the motor tasks used by neurologist to assess the severity of the impairment on lower limbs as defined in the Unified Parkinson's Disease Rating Scale (UPDRS). In particular, attention has been focused on the characterization of the Leg Agility task through a set of kinematics variables, able to capture all those features of the movement (such as amplitude, speed, rhythm, and typical anomalies as hesitation, freezing and interruptions) implicitly considered by neurologists during clinical evaluation; the correlation with the UPDRS neurological score and the automatic evaluation of the patient's performance in the perspective of a "daily and at-home" monitoring. An approach based on optical devices and Computer Vision algorithms has been developed for the acquisition and the tracking of body movements; while some machine learning methods, in particular supervised classifiers, have been considered for the objective and automatic evaluation of the characteristics of the movement. A detailed description of the acquisition system and methodology will be presented, as well as the results on a cohort of patients affected by Parkinson's disease, with different impairment on lower limbs and postural severity.
A vision-based characterization of Leg Agility task in Parkinson's Disease for the automatic assessment of lower limbs impairment
Claudia Ferraris;Roberto Nerino
2018
Abstract
Impairment of lower limbs is one of the most evident features in Parkinson's Disease (PD) that represents a heavy burden on the quality of life of people affected by this chronic and neurodegenerative disease. In fact, in PD there is a progressive and general degeneration of the motor control, in which serious problems involve lower limbs, for example during gait, characterized by episodes of freezing of the gait (FOG), bradykinesia or festination; but also disorders of balance, asymmetric postural attitude and postural instability. All these typical manifestations contribute to increase the risk of falls, consequently hindering the execution of the simplest activities in daily living. This technical report is related to Leg Agility, one of the motor tasks used by neurologist to assess the severity of the impairment on lower limbs as defined in the Unified Parkinson's Disease Rating Scale (UPDRS). In particular, attention has been focused on the characterization of the Leg Agility task through a set of kinematics variables, able to capture all those features of the movement (such as amplitude, speed, rhythm, and typical anomalies as hesitation, freezing and interruptions) implicitly considered by neurologists during clinical evaluation; the correlation with the UPDRS neurological score and the automatic evaluation of the patient's performance in the perspective of a "daily and at-home" monitoring. An approach based on optical devices and Computer Vision algorithms has been developed for the acquisition and the tracking of body movements; while some machine learning methods, in particular supervised classifiers, have been considered for the objective and automatic evaluation of the characteristics of the movement. A detailed description of the acquisition system and methodology will be presented, as well as the results on a cohort of patients affected by Parkinson's disease, with different impairment on lower limbs and postural severity.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
