Extract stance-phase landmarks from a time-normalised vertical GRF curve

From a vertical ground-reaction-force curve time-normalised to the stance phase (0-100 %), reads the three classic landmarks of the double-humped stance profile: the loading (weight-acceptance) peak, the mid-stance trough, and the push-off peak. Applied to a body-weight-normalised curve, the landmarks are directly comparable across subjects of different mass, and the peak-trough-peak pattern is the mechanical signature clinical gait analyses contrast between cohorts (e.g. reduced weight-acceptance modulation in pathological gait: lower peaks, a raised trough).

Usage

grfLandmarks(
  grf,
  loading_window = 5:40,
  midstance_window = 30:60,
  pushoff_window = 45:90
)

Arguments

grf

Numeric vector: a vertical GRF curve time-normalised to the stance phase (e.g. 101 points spanning 0-100 %). Usually in body-weight units so the landmarks are subject-mass independent. A matrix (curves x time) is also accepted, in which case the landmarks are read from the ensemble-mean curve (colMeans) — the group-mean landmarks of a cohort of stance curves.

loading_window, midstance_window, pushoff_window

Integer index ranges (into grf) in which to read the loading peak (a maximum), the mid-stance trough (a minimum) and the push-off peak (a maximum). Defaults target a 101-point stance curve; widen them for populations with atypical timing. The windows deliberately overlap, so the landmarks are found in temporal order: the trough is sought strictly after the loading peak and the push-off peak strictly after the trough, guaranteeing peak1 < trough < peak2. A landmark that lands on the upper edge of its window raises a warning, since the true extremum may lie beyond it.

Value

A named numeric vector c(peak1, trough, peak2) — the loading peak, the mid-stance trough and the push-off peak, in the units of grf.

References

Winter DA (2009). "Biomechanics and Motor Control of Human Movement." 4th ed. John Wiley & Sons.

Horsak B, Slijepcevic D, Raberger AM, Schwab C, Worisch M, Zeppelzauer M (2020). "GaitRec, a large-scale ground reaction force dataset of healthy and impaired gait." Scientific Data 7:143.

See also

filterGRF() to band-limit the force trace, computeImpulse() and computeLoadingRate() for other GRF-derived measures.

Examples

# a synthetic double-humped stance curve (101 pts, body-weight units)
g <- 1.1 * sin(seq(0, pi, length.out = 101))^0.5
g[46:56] <- g[46:56] * 0.68
grfLandmarks(g)
#>     peak1    trough     peak2 
#> 1.0669891 0.7433812 1.0902145