Premium
Objective detection of subtle freezing of gait episodes in Parkinson's disease
Author(s) -
Delval Arnaud,
Snijders Anke H.,
Weerdesteyn Vivian,
Duysens Jacques E.,
Defebvre Luc,
Giladi Nir,
Bloem Bastiaan R.
Publication year - 2010
Publication title -
movement disorders
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.352
H-Index - 198
eISSN - 1531-8257
pISSN - 0885-3185
DOI - 10.1002/mds.23159
Subject(s) - gait , parkinson's disease , physical medicine and rehabilitation , treadmill , gait analysis , psychology , medicine , physical therapy , audiology , disease
Freezing of gait (FOG) is a clinically defined phenomenon of Parkinson's disease (PD). Recent evidence suggests that subtle FOG episodes can be elicited in a gait laboratory using suddenly appearing obstacles during treadmill walking. We evaluated which quantitative gait parameters identify such subtle FOG episodes. We included 10 PD patients with FOG, 10 PD patients without FOG, and 10 controls. Subjects walked on a motorized treadmill while avoiding unexpectedly appearing obstacles. Treadmill walking was videotaped, and FOG episodes were identified by two independent experts. Gait was also analyzed using detailed kinematics. Knee joint signals were processed using time–frequency analysis with combinations of sliding fast Fourier transform and wavelets transform. Twenty FOG episodes occurred during treadmill walking in 5 patients (all with clinically certified FOG), predominantly in relation to obstacle avoidance. FOG was brief when it occurred just before or after obstacle crossing and was characterized by short, rapid steps. Frequency analysis showed a typical qualitative pattern: before the FOG episode an increase in dominant frequency in the 0 to 3 Hz band (festination), followed by decreased power in 0 to 3 Hz band and an increased power in the 3 to 8 Hz band during the FOG episode. This pattern led to an increased FOG index as a qualitative measure. These approaches detected even very brief FOG with acceptable sensitivity (75–83%) and specificity (>95%). We conclude that time–frequency analysis is an appropriate approach to detect brief and subtle FOG episodes. Future work will need to decide whether this approach can support or even replace expert clinical opinion. © 2010 Movement Disorder Society