Abstrakt
Turning is an essential part of human movement. Turning manoeuvres are affected by age, neurological disorders, or frailty.
Analyses of the walking turn and its alterations could provide valuable information about functional independence. Most studies involving wearable sensors quantify the turn by descriptive statistical values such as the mean or maximum of the signal.
Along with growing interest in walking turn analysis, new parameters should be proposed. From statistics we adapted a parameter, referred to as waveform skewness, that describes the shape of the 180 degrees turn signal.
This parameter is then compared to established ones and an intraclass-correlation is calculated. The mutual relationship between the proposed parameter and established ones is investigated via correlation.
In addition, the effect of different circumstances (temporal alignment, signal scaling and time shift) to the proposed parameter is quantified. Waveform skewness showed a moderate intra-class correlation and a high correlation with the signal peak value.
The results showed that waveform skewness is not sensitive to the time shift but is sensitive to signal scaling and temporal alignment. Comparing the waveform skewness between the two subject groups revealed significant differences between Parkinson disease patients and the control group.
Quantitative assessment of the 180 degrees turn may allow for more objective and sensitive determinations of movement disorders and pathologies.