VISUAL INTELLIGENCE FOR MEDICINE
The fetal heart rate is the most-watched signal in labor and the hardest to read consistently. The same tracing draws different conclusions from different clinicians, and from the same clinician on a different day. Footmark turns a tracing into objective, calibrated, fully auditable measurements, so the read no longer depends on who is looking — and it lands in the standards you already read by, NICHD and FIGO. We are building this in the service of the people who carry these decisions, beginning with the care of children.
Tracing analysis
| Time | Feature | Conf. |
|---|---|---|
| 14:02 | Baseline 140 bpm | 0.98 |
| 14:18 | Minimal variability | 0.94 |
| 14:31 | Late deceleration | 0.97 |
| 14:44 | Recurrent late decelerations | 0.96 |
The same measurements, scored under both standards — with any disagreement flagged.
Recurrent late decelerations from 14:31 with minimal variability. Every value is measured off the grid and timestamped to its source.
<40%
Agreement among board-certified OB specialists reading the same fetal heart rate tracing.
THE PROBLEM
Interpreting a fetal heart rate tracing is subjective, and the field has known it for decades. Most of what a clinician sees in labor falls into Category II, the broad middle of the three-tier system. It covers the majority of tracings and carries no agreed management. Two careful clinicians, or two shifts, can read the same record and reach different conclusions.
Forty years of continuous monitoring has not resolved this. The limiting factor has never been the absence of a tool. It is the consistency and the specificity of how the signal is read.
HOW IT WORKS
Footmark reads the tracing the way a careful colleague would, and reads it the same way every time. The night shift and the day shift see the same assessment of the same record.
Measured against the grid, not estimated by eye.
It surfaces the subtle changes that hide in a long labor — drifting variability, a deceleration pattern building over an hour, and a signal that may be the mother's pulse rather than the baby's.
Including quantitative measures the eye cannot estimate, such as total deceleration area.
Every finding shows its reasoning and points to the exact place on the tracing it came from. You can check it, agree or disagree, and trust it because nothing is hidden.
A provenance record for every value, end to end.
Footmark reports its read in the frameworks your unit already uses — NICHD and FIGO — not a category of its own. It scores the same objective measurements under both, and when the two disagree on a record, it says so, because that disagreement is exactly where a closer read is warranted.
Both standards, from one set of measurements — and where they conflict, you see it.
THE STANDARD WE HOLD
The same tracing produces the same measurements, whoever runs it. The read carries no opinion and no stake in the outcome.
Every number is anchored to the real gridlines and the printed paper speed. Where the record is ambiguous, the system reports the uncertainty rather than guessing.
Footmark measures the tracing and hands over the numbers. Every clinical judgment, and the responsibility for it, stays with the clinician.
WHAT IT MEASURES
Footmark reports the full set of features a clinician reads, plus quantitative measures a categorical read leaves out.
Baseline rate and baseline variability, measured over the standard window.
Accelerations and decelerations, classified by type and timed against the grid.
Total deceleration area and related continuous measures linked in the literature to fetal acidemia.
Signal-quality checks, including tracings consistent with maternal-heart-rate contamination, flagged for review.
A timestamped provenance record for every value, from the pixel to the number.
Both frameworks, side by side: the tracing is scored under NICHD and FIGO from the same objective measurements, with any disagreement between the two surfaced for the clinician.
One set of measurements, scored under both NICHD and FIGO, with any disagreement surfaced for the clinician.
THE SCIENCE
The three-tier system was built to help readers agree with one another, not to predict which fetuses are at risk. Most labors sit in its broad middle. We think the next step is not more categories. It is objective, quantitative measurement, validated against real outcomes.
The hard part has never been the algorithm. It is the data — large, diverse, outcome-linked tracings, measured objectively. Most fetal-monitoring AI is trained on the same small public dataset and behaves like a black box. Footmark's technology produces the data the field is missing, and reports a read that shows its work. We are building that evidence base, and we are honest about the bar: any new measure has to predict outcomes and hold up beyond a single dataset.
If defining the next generation of fetal monitoring on real evidence is a problem you want your name on, we should talk.
Get involvedCOMMON QUESTIONS
GET INVOLVED
Footmark is early, and the medicine has to lead. If you read these tracings, train the people who do, or study them, we would value your eyes on the work.