Translating published metrics to what ForceIQ shows

When a paper reports RFD at 50 ms or CV at 20% MVC or Yank in N/s, how does that map to what the platform shows on the metric tile? A practical bridge for clinicians reading the literature, with the most common units, window conventions, and normalization differences spelled out.

Updated May 15, 2026

The published literature on isometric force assessment is heterogeneous on units, time windows, and normalization conventions. Two papers can both report "RFD" and mean different things by it; two papers can both report "CV" and have computed it across different windows; one paper's "Yank" is comparable across patients and another paper's "Yank" is not. The platform's metric tiles use a specific set of conventions, and the conventions are deliberate, but the gap between "what ForceIQ shows" and "what a paper you're reading reports" is one of the most common sources of confusion when clinicians try to apply literature thresholds to their own patient data.

This article maps the platform's conventions to the most common ways the published work expresses the same metrics, so a paper's reported value can be translated to (or compared against) what shows on the screen.

Rate of force development (RFD)

What the platform reports. Two time-locked slopes per trial: RTD Early (0 to 75 ms) and RTD Late (100 to 200 ms). Both computed as two-point interpolated secants between the force at the window's start time and the force at the window's end time, in Nm/s when torque conversion is available or N/s otherwise. No peak RFD tile.

How the literature commonly reports it. Three common variants, plus several less-common ones:

  • Window-locked RFD at specific time points (RFD 0-50 ms, RFD 0-100 ms, RFD 0-150 ms, RFD 0-200 ms). This is the closest match to what the platform reports. RFD 0-75 ms is the platform's RTD Early; "RFD 0-100 ms" or "RFD 0-150 ms" in a paper sits in between the platform's two windows but is mechanistically dominated by the contractile machinery rather than neural drive after about 75 ms, so it's not directly comparable to either tile (Del Vecchio 2022; D'Emanuele et al. 2023).
  • Peak RFD. The instantaneous maximum of the first derivative of force. The platform does not surface a peak RFD tile, because peak RFD is dominated by maximum voluntary force and may not add information beyond peak itself (D'Emanuele et al. 2023). When a paper reports "peak RFD" as the primary RFD outcome, treat it as a peak-strength variable rather than a separate "speed of force production" variable.
  • RFD as average across a window (RFD_avg from 0 to 200 ms, say). Same window-bounded approach as the platform's tiles, but typically computed as (force at end) divided by (window duration), without interpolating the endpoints to land exactly on the target times. At high sample rates this is similar to the platform's value; at lower sample rates it can differ by 5 to 15 percent depending on where the nearest samples fell.

Practical translation. A paper reporting "RFD 0-75 ms" or "early-phase RFD ≤ 75 ms" matches the platform's RTD Early. A paper reporting "RFD 100-200 ms" or "late-phase RFD" matches RTD Late. Most other RFD windows in the literature are intermediate and not directly comparable to either tile. If you need to compare against a paper that reports RFD at a specific intermediate window (say 50 to 150 ms), the platform's two-window output gives you bracketing values rather than an exact match.

Coefficient of variation (CV)

What the platform reports. CV across the plateau window, computed as (SD / mean) × 100. Plateau detection is automatic; the window starts after the rise to peak and ends before the release, typically a sustained-force region of 1 to 3 seconds. Reported as a dimensionless percent.

How the literature commonly reports it. Several axes vary across papers:

  • The target force. Many force-control papers report CV at specific target intensities (5%, 10%, 20%, 30%, 50% MVC), not on maximum-effort plateaus. CV is force-dependent: it falls as the target rises (Enoka & Farina 2021). A paper's CV at 10% MVC is not directly comparable to the platform's CV from an MVIC plateau. To compare against literature CV at submaximal targets, use an IsoForge submaximal-hold session at the matching percentage; the platform stores those summaries on the patient's record. See setting submaximal targets from a recent MVIC.
  • The plateau duration. Some papers report CV over the first 1 second of a hold, others over the middle 5 seconds, others over the full 30 seconds. The platform's plateau detector typically produces a window in the 1 to 3 second range on an MVIC trial. CV is reasonably stable across plateau durations as long as the patient hasn't fatigued; on long holds with fatigue, the CV reflects the fatigue trajectory.
  • The filter applied. Some papers low-pass filter aggressively before computing CV, others compute on raw signal. The platform's CV is computed on the same filtered signal it uses for plateau detection. Heavy filtering reduces CV; raw-signal CV is higher.

Practical translation. When a paper reports CV at a percent-MVC target, run an IsoForge submaximal hold at the matching target on the same patient and read the training summary rather than the assessment session. When a paper reports CV on MVIC plateaus with comparable filter and window conventions, the platform's CV is directly comparable.

Yank (RMS of dF/dt during the plateau)

What the platform reports. Yank as RMS(dF/dt) divided by peak force, in s⁻¹ (dimensionless rate). The amplitude normalization is deliberate; it makes Yank comparable across patients and target forces.

How the literature commonly reports it. Yank is a recent enough metric that the conventions are still being established. Two variants in the published work:

  • Absolute RMS(dF/dt) in N/s. This is the variant proposed by Yacoubi & Christou (2024). Without normalization, the value scales with target force and with movement amplitude, which makes it useful within a single subject and protocol but not comparable across patients or sessions at different targets.
  • Normalized RMS(dF/dt). Multiple normalizations are possible. The platform uses amplitude normalization to peak force per Sherman et al. (2024), which produces the dimensionless s⁻¹ values on the metric tile. Other normalizations from the jerk literature (dimensionless jerk, log-dimensionless jerk, spectral arc length) are also defensible but produce values on different scales (Sherman et al. 2024).

Practical translation. A paper reporting absolute Yank in N/s on an MVIC plateau is comparable to the platform's Yank only after the absolute value is divided by the paper's reported peak force, which most papers don't make convenient. A paper reporting Yank in s⁻¹ may or may not be comparable depending on which normalization they used; the methods section is the only place to check.

For now, the safest comparison axis on Yank is within-patient longitudinal change. The platform's Yank is internally consistent across sessions, so a patient's Yank trajectory across visits is interpretable even when the absolute value is hard to map to specific published thresholds.

Peak force and peak torque

What the platform reports. Peak force in N. If patient height is set, also peak torque in Nm using the estimated moment arm. If bodyweight is set, also normalized to mass (N/kg or Nm/kg).

How the literature commonly reports it. Knee-extension and other joint-isolated isometric work is usually reported in Nm/kg (mass-normalized torque) at a specific joint angle. Common reporting:

  • Peak torque in Nm/kg at 60° knee flexion is the dominant convention in post-ACLR literature. The platform's torque value at the matching protocol is directly comparable.
  • Peak force in N without torque conversion. Less common in clinical literature; more common in research protocols that don't measure moment arm. The platform's force value is directly comparable.
  • Peak force in lbs or kg. Convert at the standard ratios: 1 lb ≈ 4.45 N, 1 kgf ≈ 9.81 N. The platform's display unit can be toggled to match a paper's units.

Practical translation. For knee extension in Nm/kg at 60°, the platform's value is directly comparable to most published normative ranges (Norris et al. 2024). For other joints or non-standard angles, comparability depends on whether the platform's moment-arm model matches the paper's measured moment arm; usually close enough for clinical interpretation but not for research-grade equivalence.

Limb Symmetry Index (LSI)

What the platform reports. LSI as the weighted average across the metric set (peak weighted highest, RTD windows second, steadiness third), with per-metric components capped at 100% before averaging. Plus per-metric LSI on the radar.

How the literature commonly reports it. The published LSI is almost always a single-metric ratio: peak strength LSI, RFD LSI, hop-test LSI. The platform's overall LSI is a weighted blend; published per-metric thresholds (a "90% LSI cutoff" in a paper) almost always refer to peak strength specifically, not to the platform's headline number.

Practical translation. When a paper reports an LSI threshold for return-to-sport (commonly 85% or 90% on peak strength), apply that threshold to the platform's per-metric peak LSI on the radar, not to the overall headline. The headline is a composite; the per-metric tiles are what map to the literature. See Limb Symmetry Index, the weighted average and what the headline hides for the deeper read.

The general rule

When in doubt, compare the platform's value to a paper's value only after checking three things: the window, the normalization, and the protocol. The window: is the time interval (for RFD) or the plateau duration (for CV / Yank) the same? The normalization: is the metric amplitude-normalized, mass-normalized, both, or neither? The protocol: was the patient doing the same kind of contraction (MVIC vs ballistic vs submaximal hold)?

When all three match, the values are comparable. When any of the three differs, the values can still be informative but require interpretation rather than direct comparison.

What to do next

  • Read the methods section before reading the results. The window, normalization, and protocol details determine whether the paper's numbers are comparable to yours.
  • For RFD, the platform's RTD Early matches most papers' RFD 0-75 ms. Other window choices are common; check the bracket.
  • For CV at submaximal targets, run an IsoForge submaximal hold, not an MVIC. Comparing assessment-session CV to literature submaximal-CV is comparing different protocols.
  • For Yank, longitudinal change is more comparable to literature than absolute value. The normalization conventions are still settling.

References

  • Del Vecchio A. Neuromechanics of the rate of force development. Exerc Sport Sci Rev. 2022. doi:10.1249/JES.0000000000000306
  • D'Emanuele S, Tarperi C, Rainoldi A, et al. Neural and contractile determinants of burst-like explosive isometric contractions of the knee extensors. Scand J Med Sci Sports. 2023;33(2):127-135. doi:10.1111/sms.14244
  • Enoka RM, Farina D. Force steadiness, from motor units to voluntary actions. Physiology (Bethesda). 2021;36(2):114-130. doi:10.1152/physiol.00027.2020
  • Yacoubi B, Christou EA. Rethinking force steadiness, a new perspective. J Appl Physiol. 2024;136(5):1260-1262. doi:10/gtvkp7
  • Sherman DA, Darendeli A, Soto O, et al. Beyond force steadiness, potential challenges in measuring smoothness of force through yank. J Appl Physiol. 2024;136(5):1266-1267. doi:10/gtvpr5
  • Norris R, Morrison S, Price A, et al. Inline dynamometry provides reliable measurements of quadriceps strength in healthy and ACL-reconstructed individuals and is a valid substitute for isometric electromechanical dynamometry following ACL reconstruction. The Knee. 2024;46:136-147. doi:10.1016/j.knee.2023.12.006
Was this helpful?