| To measure | sensors/hardware | readings needed | calculation | method | notes | how often to measure |

| ------------------------------------ | --------------------------------------------------------------------------------------------- | ------------------- | ---------------------------------------------------------------------------- | -------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------- |

| Heart Rate | green LED (~530nm), photodiode
accelerometer for noise, band pass filter | | | PPG | filter for 0.5-4 Hz (≈ 30–240 bpm) | |

| Heart Rate Variability | | HR reading | changes in time between heart beats
RMSSD
measured in milliseconds | PPG | measure during sleep,
need continuous HR data for a period of measurement | during sleep |

| SpO2 (oxygen saturation) | red (660nm), IR (940nm) LED, photodiode | | see SpO2 note | | how to account for differences in skin tone?
only works on radial artery
| |

| Anemia | red, IR, green, blue LED, photodiode | | compare rates of absorption of hemoglobin at different wavelengths | | *screening
high SpO2 != no anemia
consistently low = flag
measure during sleep | |

| Sleep apnea | | SpO2, RR | calculate drops in oxygen during sleep
see Sleep Apnea | | *screening | |

| Chronic stress | skin temp, EDA electrodes (electrodermal activity)/sweat sensing, | HRV, resting HR, RR | | | lower RMSSD = generally more stress (over time) | |

| Arrhythmia, Bradycardia, Tachycardia | accelerometer | HR | irregular pulse algorithm | PPG + accel | | probably during sleep? |

| Hypoxemia (low blood oxygen) | barometer (for altitude filtering) | SpO2 | low SpO2 over period of time: <92% low and <88% high risk | | | |

| Respiratory Rate (RR) | green LED, photodiode, accelerometer
*same as HR, but looks at different part of waveform | | | PPG + analysis | lower RR = better overall fitness
higher overall = fatigue/stress
combine with accel data to reduce motion noise
normal ~12-20 breaths/min (adult)
measure at night
| during sleep |

| VO2 Max | | HR | find an algorithm | | | |

| Sleep | accelerometer for motion | HR, SpO2, RR, HRV | | | determine sleep phase, length of sleep | |

| Hydration | skin conductivity/sweat sensing | | | | high skin conductivity = more salt/electrolytes = less water = less hydration | |

other hardware:

  • gps?

  • SOM

    • does this interface with sensors directly?

    • averaging/smoothing functions

      • lots of things are over time measurements
  • filtering (mentioned above)

    • esp important for HR

    • other things that require extensive filtering - i noted to measure during sleep. unsure though

  • skin temp correlates with menstrual cycle & possibly other hormones - def worth

  • ambient light sensor to tune LED readings

  • use IMU for exercise/activity tracking

    • did not get to it in the table

thoughts:

  • option to turn on/off these readings

  • term for screenings - “flags” or “risk alerts”

  • children vs adults?