The Limb Symmetry Index is the most cited single number in return-to-sport decision-making. It compresses every metric on a bilateral assessment into one ratio, which is convenient for charting and for thresholds, but the convenience hides three things worth understanding: the weighted average that produces the headline, the per-component cap that prevents a single inflated metric from rescuing a deficient one, and the "lowest metric" line that renders beside the weighted headline in the expanded view so a green overall cannot hide a red component.
The headline and what's in it
LSI is the ratio of the involved-side value to the uninvolved-side value, expressed as a percentage. 100% is symmetric, below 100% means the involved side is weaker (or less steady, depending on the metric). The platform computes a per-metric LSI for every metric on the bilateral assessment (Strength, Explosiveness, Build-up, Drift from peak, Steadiness and Smoothness, or peak, RTD Early, RTD Late, nRMSE, CV and Yank) and then averages them into the weighted overall, the expanded view's headline.
The average is weighted, not equal. Peak carries more weight than the RTD windows; the RTD windows carry more weight than the Control metrics. The weighting reflects the literature's emphasis: peak symmetry is the most consistently studied predictor of re-injury risk; explosive strength is the second-tier criterion; steadiness is a relatively new entrant. The exact weights are set by the platform; they're tuned to roughly match how clinicians describe their own decision-making in published return-to-sport criteria sets.
LSI
mean of 3v3 trials
- Current
- Goal (100%)
Two headlines: Strength LSI and the weighted LSI
The dashboard shows each patient in one of two views, and the view decides which LSI leads. Every patient starts in the simplified view, which leads with Strength LSI: peak symmetry on its own, the involved side's mean peak as a percentage of the uninvolved side's. It is not capped, so a value above 100% means the involved side produced more. The simplified view also regroups the six per-metric LSIs into three: Strength (peak), Speed (Explosiveness and Build-up, the RTD Early and RTD Late windows) and Control (Drift from peak, Steadiness and Smoothness, which are nRMSE, CV and Yank), each a weighted average capped at 100% per component, exactly as the overall is.
The expanded view leads with the weighted LSI the rest of this article describes, with the lowest metric beneath it and the six-spoke radar.
Both views read the same assessment and the same per-metric LSIs; they differ in which number leads. Early in rehab, peak symmetry is usually the question, and a weighted overall that folds in Speed and Control, which recover later, reads lower than the strength picture alone. Whichever view a patient is on also sets the LSI badge and the 90% Flagged rule on the clinic overview, so a patient is flagged on the number you see when you open them. See reading the dashboard for the panel itself.
Why components are capped at 100% before averaging
A patient who happens to be less variable on the involved side than the uninvolved side will produce Drift from peak, Steadiness, or Smoothness LSIs (nRMSE, CV, or Yank) that read above 100% (the involved side is steadier). Without a cap, that 110% Control symmetry would pull the weighted average up and partially offset, say, a 75% peak symmetry. The resulting overall would suggest the patient is closer to recovered than they actually are.
To prevent that, every per-metric component is capped at 100% before contributing to the weighted average. Above-symmetric values are still surfaced on the radar (so you can see that the involved side was steadier, which is sometimes clinically meaningful), but they cannot rescue the headline. A patient at 75% peak symmetry and 110% Smoothness (Yank) symmetry has an overall pulled down by peak; the Smoothness value sits at its raw 110% on the radar but contributes a capped 100% to the average.
This is a small mechanical detail that matters because the alternative reads badly in practice. Clinicians making return-to-sport decisions on a "92% LSI" need to know that number reflects the patient's deficits, not the patient's deficits offset by an irrelevant strength on the involved side.
The "lowest metric" surface
A weighted average can look closer to recovered than the patient is. 93% overall sounds nearly there. But if that patient has 68% on Explosiveness (RTD Early) and is symmetric on the other five metrics, the patient is not actually at 93% on the thing that matters most for cutting, jumping, or landing. The headline number is technically correct but clinically incomplete.
The LSI panel surfaces the lowest individual per-metric LSI on a line directly underneath the headline, with the same color-tier styling (red below 70%, amber 70 to 85%, yellow-green 85 to 95%, green at 95% and above). On a patient with 93% overall and 68% on RTD Early, that line names Explosiveness (RTD Early) at 68%, in red. You see both numbers side by side: the yellow-green headline and the red component pulling against it.
The clinical read on that pattern is usually that the patient is strong but slow to fire. The peak deficit has resolved, but the explosive phase hasn't. It is a familiar pattern after ACL reconstruction: between-limb asymmetry in RFD can persist for years after surgery, and in the pooled data early-phase knee extensor RFD sits about a standard deviation below the uninvolved limb (standardized mean difference −1.07; Turpeinen et al. 2020). The line on the panel exists so that pattern doesn't get missed by glance-reading the overall.
The radar shape
The radar plots all six per-metric LSI values on the same scale. The shape tells you what kind of asymmetry the patient has, not just how much. Two patients with the same 85% overall LSI can have very different radars:
- A uniform radar: every metric sits around 85%. The patient has a global strength deficit on the involved side. Rehab focuses on building capacity across the board.
- A peaked radar: Strength and Speed are recovered (95%+), but the Control metrics drag the average down (70% on Drift from peak, Steadiness, or Smoothness). The patient is strong but not controlled. Rehab focuses on motor control work.
- A valley radar: Strength is recovered but Explosiveness and Build-up are sub-85%. The patient is strong but slow. Rehab focuses on explosive training.
The radar is the part of the panel most worth reading first. The headline is what you write in the chart note; the radar is what you build the next phase of rehab against.
Try it in the Dynamometry Lab: compare a real patient's two knees metric by metric, and see why the Control metrics are inverted and each metric capped at 100 %.
Control in LSI specifically
The three Control metrics in the average (Drift from peak, Steadiness and Smoothness: nRMSE, CV and Yank) are an active area of measurement research, and treating them as ratios of involved-to-uninvolved is not without subtlety. Force steadiness depends on muscle size, motor unit population, and target force, with small muscles producing lower CV at the same percent-MVC than large muscles (Enoka & Farina 2021). For within-patient LSI between the same patient's two legs, those confounders are largely controlled (same person, same target percentage). For cross-patient comparisons of Control LSIs, more caution is warranted.
The post-ACL literature also suggests that nonlinear analyses of force control detect more and larger asymmetries than the linear coefficient-of-variation methods (Schwartz et al. 2025). The platform reports CV (Steadiness) as the linear baseline; Yank (Smoothness) captures a complementary axis (smoothness, not variability) that the linear methods miss. See Control, reading drift from peak, steadiness, and smoothness together for the full picture.
What to do next
- Read the radar first, the headline second, the lowest-metric line third. All three in one scan; the radar tells you the shape, the headline tells you the magnitude, the lowest metric tells you the limiting factor. In the simplified view the same order holds with the Strength / Speed / Control triangle and the lowest measure named under Speed and Control.
- For return-to-sport decisions, look at all six per-metric LSIs against your published thresholds, not just the headline against a 90% rule of thumb. The literature's thresholds are typically per-metric, not overall.
- A green overall with a red lowest metric is the most informative pattern on the panel. It's what the panel was designed to surface.
References
- Turpeinen JT, Freitas TT, Rubio-Arias JÁ, et al. Contractile rate of force development after anterior cruciate ligament reconstruction, a comprehensive review and meta-analysis. Scand J Med Sci Sports. 2020;30(9):1572-1585. doi:10.1111/sms.13733
- Schwartz AL, Koohestani M, Sherman DA, et al. Knee extensor and flexor force control after ACL injury and reconstruction, a systematic review and meta-analysis. Med Sci Sports Exerc. 2025;57(2):238. doi:10.1249/MSS.0000000000003574
- 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