The validated results on this site come from one demonstration case: elite women's ice hockey. Below is exploratory thinking about other sports and settings — offered as hypotheses worth stating plainly, not as findings.
Structural waveform shape, rotational–translational coupling, and spectral content are properties of any recorded head-motion event, whatever the sport. What's untested is whether the specific measures and thresholds calibrated on ice-hockey data transfer to other impact mechanisms, equipment, and sensor placements — each would need its own validation, exactly as this framework required for ice hockey.
Long before wearable sensors or modern imaging, clinicians recognized that repeated blows to the head could cause lasting neurological damage. In 1928, pathologist Harrison Martland described a progressive syndrome in boxers — popularly known at the time as being "punch drunk" — later formalized as dementia pugilistica: tremor, slowed movement, and cognitive decline emerging years after a career had ended. By the late 1940s the condition was already being discussed in the medical literature under the name chronic traumatic encephalopathy.
The same underlying pathology — under that modern name, CTE — has since been documented in other contact and collision sports, including American football, where it has been the subject of extensive neuropathological research over the past two decades, including at the professional (NFL) level. CTE can currently only be diagnosed after death, through post-mortem examination of brain tissue — there is no validated test for it in a living athlete. That diagnostic limitation is part of why an objective, in-vivo description of actual mechanical exposure — not just an average magnitude, but the structure of each impact over a career — is worth pursuing.
Sources: References.
Purposeful heading is a directed, anticipated, sustained contact with the ball — mechanically distinct from an incidental head-to-head or head-to-ground collision, even at similar peak magnitude. If that distinction holds, it should appear as a difference in waveform duration, symmetry, and rotational coupling — the structural and coupling domains — the same way on-ice contact separated from off-ice disturbance in the ice-hockey data. Separating anticipated heading from incidental contact this way has not been tested.
American football produces a far higher volume of repetitive, moderate-magnitude head contact than most contact sports — offensive and defensive linemen alone can sustain hundreds of head impacts across a single season, mostly from blocking rather than open-field tackling. It is also the sport where CTE has been most extensively studied neuropathologically, including at the NFL level. Current exposure monitoring in football, like elsewhere, is built around peak magnitude and impact counts. If the structural and spectral domains can distinguish a blocking-line collision from an open-field tackle from a helmet-to-helmet impact — categories with overlapping peak magnitudes but plausibly different waveform shape and frequency content — that would be a more detailed exposure picture than counts alone. Untested — no American football data has been analyzed with this framework.
Glove or fist contact is typically short-duration and rigid, plausibly showing high-frequency, narrow-band spectral content — versus grappling, clinch, or positional contact, which is likely to be more sustained and lower-frequency. If so, the spectral and structural domains could in principle separate strike types from incidental contact within a bout, the way spectral content separated genuine impacts from artefacts in the ice-hockey artefact-detection analysis. This has not been tested in combat-sport data.
A fall onto ice is a sustained, body-mass-mediated contact, mechanically closer to the "on-ice" gameplay pattern than to a brief sensor disturbance. Across a season, repeated falls could plausibly be tracked as a distinct exposure category using structural and spectral measures, separate from the sport's routine, low-magnitude motion. Untested — no figure-skating data has been analyzed with this framework.
Because peak magnitude alone doesn't capture mechanical structure, two athletes with similar peak-acceleration exposure could have very different structural profiles — shaped by head mass, neck strength, or preparatory muscle activation. In principle, each athlete could accumulate an individual baseline region within the feature space, and deviation from that personal baseline could be a more sensitive exposure signal than deviation from a population norm. This individual-baseline approach has not been implemented or tested longitudinally, in ice hockey or any other sport.
Brain deformation depends on the shape and duration of the mechanical input, not only its amplitude — a relationship well established in finite-element and reduced-order brain models. The waveform characteristics IFS already computes (duration, rise shape, frequency content) are the same kind of inputs those models use. Whether IFS features have any predictive value for brain-strain modelling is, at this stage, an open question rather than a claim.
For any of the above: sensor data recorded with a validated, sport-appropriate placement; independent, non-sensor-derived ground-truth labels for the event categories of interest (as Wisehockey player-tracking provided for on-ice/off-ice in the ice-hockey work); and, ideally, an outcome measure — injury, exposure severity, or a validated proxy — to test against. None of that currently exists for the sports listed above.
If you work with a sport or setting where this kind of validation could be run, we'd like to hear from you.