Reading IsoForge training summaries

An IsoForge session saves a training summary, not the full clinical metric suite. The two surfaces are kept separate deliberately because they answer different questions. Assessment metrics tell you where the patient is. Training summaries tell you how the training is going. This walks through what each summary contains, what it does and doesn't replace, and how to use the two together across a rehab block.

Updated May 13, 2026

When you end an IsoForge session, the platform saves a training summary. When you end an assessment session (MVIC, RTD/ballistic, submaximal hold via the assessment workflow), the platform saves the full clinical metric suite. These are intentionally different surfaces, and confusing one for the other is the most common interpretation error in the IsoForge workflow.

The simple version: assessment sessions answer "where is the patient." Training summaries answer "how is the training going." Both matter; neither replaces the other.

What an IsoForge training summary contains

The fields vary slightly by game mode, but the common shape:

  • Set count and total active time. How many sets the patient completed and how much cumulative time they spent producing force.
  • Mean force across the session. Average force value during active periods, in the unit you've selected (N, lbs, or kg).
  • Target adherence. For Hold Steady (the sustained-hold mode), this is time-in-zone (the percentage of active time the patient held inside the target band). For Make Waves (sinusoidal tracking), it's tracking error (RMS distance between the patient's force and the moving target). For Take Flight (the impulse/RFD mode), it's gates cleared vs gates attempted. For Rapid Fire (chirp pulse trains), it's peak-amplitude-match rate and peak-timing-match rate.
  • Rep count or impulse count. For the impulse modes (Take Flight, Rapid Fire), the count of individual force events produced.
  • Target intensity. The percent-MVIC the session was run at (for Hold Steady and Make Waves), or the difficulty setting (for Take Flight and Rapid Fire).

What the summary does not contain: peak force, RTD windows, plateau steadiness metrics (nRMSE, CV, Yank), or LSI. These are deliberately not computed on IsoForge sessions.

Why the clinical metric suite is not computed on IsoForge sessions

Three reasons:

The contraction shapes don't fit. The clinical metrics are defined for specific contraction shapes. Peak and RTD assume an MVIC-style explosive rise to a maximum. nRMSE and CV assume a sustained plateau at the patient's maximum. Yank assumes the same plateau. An IsoForge session produces submaximal holds (Hold Steady), oscillating force profiles (Make Waves), or sequences of small impulses (Take Flight, Rapid Fire). None of these shapes match the clinical metric definitions, so computing the clinical metrics on them produces meaningless numbers that look superficially clinical.

The patient isn't trying to hit peak. Even on the most explosive IsoForge mode (Take Flight), the patient is producing dozens of small impulses, not three maximum-effort trials. Their peak force on a Take Flight set is whatever the highest impulse happened to reach, which is not the same as their MVIC. A "peak" computed from an IsoForge session and a peak computed from an assessment session can differ by 30 to 50% in either direction, and the differences are about the protocol, not the patient.

The clinical metrics need to stay comparable across visits. If MVIC peak from a Tuesday assessment and Take Flight highest-impulse from a Thursday training session both showed up as "peak" on the dashboard, the longitudinal track would be incoherent. The platform keeps the two surfaces separate so trend lines stay readable.

How to read the training summary across a block

The training summary is most useful as a within-block trajectory tool. Single-session values are noisy; the trajectory across a 6 to 12 session block is what tells you whether the training is working.

Time-in-zone trend (Hold Steady, Make Waves). If time-in-zone is climbing across sessions at the same target intensity, the patient is improving at the task. If it's flat across multiple sessions, the current intensity isn't driving further adaptation; consider progressing intensity per setting submaximal targets.

Tracking error trend (Make Waves). Falling tracking error means the patient is following the target more precisely. Watch for plateau patterns; tracking error often improves quickly at slow sine frequencies and stalls when frequency is increased.

Gates cleared (Take Flight). Both the absolute count and the success rate matter. A patient clearing 20 of 25 gates is doing better than the same patient clearing 30 of 50 gates, even though the count is lower; the rate captures the consistency of the impulse production.

Peak-amplitude-match and peak-timing-match (Rapid Fire). Amplitude match usually improves first; timing match lags. Patients can produce the right size pulse before they can produce it at the right moment.

How the assessment side reads the training side

Assessment sessions during a training block answer a different question: did the training drive a measurable adaptation on the clinical metrics. The expectation:

  • A Hold Steady block focused on steadiness should produce improvements in the assessment-side nRMSE and CV on the matching joint, not in peak. If steadiness LSI rises across the block, the training worked. If peak rises, that's a bonus, but probably from a different mechanism.
  • A Take Flight block focused on explosive output should produce improvements in the assessment-side RTD Early (and possibly RTD Late). If peak rises during a Take Flight block, the explanation is usually that the patient was capacity-limited on peak going in and the increased neural drive carried over.
  • A Make Waves block focused on motor-control precision should produce improvements in CV and Yank, with Yank often more responsive than CV because Yank is sensitive to smoothness specifically.

When the training-side trajectory and the assessment-side trajectory move together, the block is working. When they diverge (training summary improving, assessment metrics flat), the training stimulus isn't transferring to the clinical reference task; reconsider the matchup between game mode and assessment deficit.

What to do next

  • Don't read training summaries as assessment data. Time-in-zone is not steadiness LSI; gates-cleared is not RTD Early. They're related, but they don't substitute.
  • Trend the summary within a block; trend the assessment metrics across blocks. Different cadences, different signals, different decisions.
  • Re-assess between blocks. The assessment is what tells you whether the IsoForge block transferred to the clinical metric you cared about.

See also choosing an IsoForge game mode for which training-summary axes to weight when picking the mode for a specific patient, and dosing IsoForge sessions across a rehab block for how the assessment cadence sits inside the training block.

Was this helpful?