A method for calculating the probability of tripping while walking.

Despite tripping being one of the frequently reported causes of falls, currently there is no method of quantifying the probability of an individual's foot contacting obstacles during gait. This paper describes a statistical modeling technique based on variability in minimum toe clearance (MTC) data during treadmill walking to estimate the probability of the toe contacting an obstacle. MTC data were calculated from two foot markers and using a 2D geometric model of the distal end of the foot. Probability of tripping (PT) was calculated by modeling and then integrating the MTC sample distribution. Results from a young male subject continuously walking for 1 hour show the MTC distribution is not normally distributed with mean=1.03 cm, S.D.=0.25 cm, skew=1.01 and kurtosis=3.47. For this distribution, PT for an unseen 0.2 cm high obstacle is calculated to be 1 in every 10,363 strides. Without skew- and kurtosis-modeling PT reduced to 1 in every 1901 strides, which highlights the importance of skew and kurtosis-modeling for PT estimation. Predicted PT is seen to increase with increasing obstacle heights (e.g. PT for an unseen 0.5 cm obstacle is 1 in 95 strides and PT for an unseen 1.0 cm obstacle is 1 in 2 strides). The method presented in this paper shows that variability in MTC data can be modeled to quantify the probability/risk of tripping on obstructions/obstacles in the travel terrain, and has the potential for wide application in the areas of falls prediction and falls minimization.