Skip to main content
A 60-second reading that passes the quality check returns these values in a VitalsResult.

Result fields

Heart rate, HRV (RMSSD) and SDNN are always present on a result. The other values are optional (nil / null) when they could not be recovered. Every value is finite. LivePreview values during the reading are provisional and for display only, so store the final VitalsResult.

Result JSON

Store results in this shape. iOS Codable writes it, Flutter toMap() returns it, and the wrappers deliver it in the completed event.
  • iOS: JSONEncoder().encode(result) writes exactly this, whatever the encoder’s date or data strategy (key strategies such as .convertToSnakeCase still apply). Decoding accepts either UUID case and capturedAt with or without fractional seconds. A result the SDK delivers decodes equal to itself.
  • Android: no serializer. id is a lower-case UUID string and capturedAt epoch milliseconds, so store the eleven properties under these keys, with capturedAt formatted as above.
  • Flutter: result.toMap() returns this map, and VitalsResult.fromMap reads it back equal.
  • React Native and Capacitor: the completed event’s result is this object, so JSON.stringify(result) stores it.

Quality tier

Every result is clean or usable. A reading the SDK cannot stand behind never becomes a result. A finished reading fails the quality check when the SDK is not confident in its HRV, when its values fall outside physiological plausibility, or when no heart rate was found. It then returns no numbers, and the next reading starts by itself (how retries work).

Breathing-rate confidence

Every result has a breathing-rate tier. It says how the rate was measured and how far to trust it. The published breathing figure comes from the camera during paced breathing. Rates from the phone’s motion (high) are not yet part of the validation, see Limits. The tier is separate from the quality tier: a clean reading can still have no breathing rate.

Validated accuracy

The figures come from the NEUROFIT validation study: 30 participants, 265 accepted readings, each a 60-second seated reading taken while wearing a Polar H10 chest strap, with the strap’s beat-to-beat timing as the reference. Half the readings were at rest and half during paced breathing at 6 breaths per minute. Three readings with frequent irregular heartbeats were excluded. The two investigators are excluded from every figure. The breathing-rate figure is camera-based. It covers 95 paced readings from 28 participants in which the chest strap confirmed the pace and the SDK returned a rate. Breathing rate outside paced breathing is an estimate, pending further validation.

Consistent across groups

No group had a reading off by 20 ms or more. Groups with one or two people describe those individuals, not the whole group. The full method and plots are in the validation report (PDF). These are interim development results. The finished system is being tested blind on new participants, and the figures will be updated as the study grows.

Real-world reliability

In everyday, unsupervised use of the NEUROFIT app, about 9 in 10 sittings where a reading started ended with a result, including automatic retries. The median time to a result was 61 seconds from the start of the reading. Source: the NEUROFIT app’s Vitals feature, more than 900 people, August to October 2026.

Limits of the validation

The accuracy figures apply to seated, still, 60-second readings. Not yet claimed:
  • Accuracy for Monk skin tones 7, 9 and 10 (no participants in the study so far).
  • Accuracy during movement.
  • Breathing rate outside paced breathing at 6 breaths per minute, or measured from the phone’s motion.
  • Readings shorter or longer than 60 seconds.
Most participants were 25 to 34 years old.

Presenting the numbers

  • Round heart rate, HRV (RMSSD) and SDNN to whole numbers, and breathing rate to one decimal.
  • Say “HRV (RMSSD)” the first time you show HRV, so users comparing with a wearable know which HRV it is.
  • Show SDNN wherever you show HRV (RMSSD).
  • When breathingRateConfidence is low, show the breathing rate as approximate and whole, as the flow does (about {value}). When it is withheld, show no breathing rate.
  • Compare a person with their own baseline, not with population tables (Outcomes best practices).
  • After cancelled(qualityCheckFailed), show the flow’s noticeQualityCheckFailed.