The first goal of the study is to investigate whether an algorithm can reliably detect Freezing of Gait (FOG) in Parkinson patients based on participant gait data generated by a pressure insole. The second goal is to investigate whether Auditive Cueing (AC) based on such a detection reduces the frequency and length of FOG episodes in those participants. The study will be conducted per Good Clinical Practice principles.
Study Type
INTERVENTIONAL
Allocation
NA
Purpose
SUPPORTIVE_CARE
Masking
NONE
Enrollment
13
Based on gait measurement, auditive cueing is generated automatically to check its impact on patients with Freezing Of Gait.
Ziekenhuis Oost-Limburg
Genk, Limburg, Belgium
The F1-score of FOG detection from the measured and classified (normal walk vs FOG) gait data compared to manually scored recorded video.
The participants perform 4 walks without AC on standardized tracks while being video-recorded. During the walks, gait data is recorded and then scored (for each time point) by the algorithm into normal walk or FOG. The video recording is scored manually according to the criteria as described in reference Gilat. M. and represents the true state of walking. The F1-score is calculated from the algorithm scoring vs the manual scoring: a True Positive is true freezing which is classified by the algorithm as freezing. A True Negative is a true normal walk which is classified as a normal walk. Similarly, a False Positive is a true normal walk which is classified as FOG and a False Negative is a true FOG which is classified as a normal walk. The same 4 walks are performed both in OFF and in ON. OFF measurement is only performed when the PI has given permission to do so. ON measurement is performed 1 hour after taking their standard medication.
Time frame: 8 walks are performed in 1 hospital visit within 4 weeks of enrollment. Total assessment time estimate is 2x 30 minutes.
The change in the number of FOG-episodes with and without AC
The participants again perform 4 walks in OFF and in ON but now with AC. The average number of freezing episodes is calculated for each participant in OFF and in ON with and without AC. The change between the average without AC and the average with AC is calculated as well as its level of significance. These changes are computed for each walk, for each individual participant across all walks and across all participants.
Time frame: 1 week after the first hospital visit (for Outcome 1).
The change in the total duration of FOG-episodes with and without AC
The participants again perform 4 walks in OFF and in ON but now with AC. The average duration of freezing episodes is calculated in the same way as their number per Outcome 2.
Time frame: within 1 to 3 weeks after the first hospital visit
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
The sensitivity of FOG detection from the measured and classified (normal walk vs FOG) gait data compared to manually scored recorded video.
Same data collection as for outcome 1. Sensitivity is calculated from the algorithm scoring vs the manual scoring.
Time frame: Within 4 weeks of enrollment
The specificity of FOG detection from the measured and classified (normal walk vs FOG) gait data compared to manually scored recorded video.
Same data collection as for outcome 1. Specificity is calculated from the algorithm scoring vs the manual scoring.
Time frame: within 1 to 3 weeks after the first hospital visit