PPG Fundamentals

What is PPG?

PPG (Photoplethysmography) is a non-invasive measurement technique that detects changes in blood volume using optical principles. By shining light onto the skin and measuring the reflected or transmitted light intensity, you can obtain important information about the cardiovascular system.

Working principle

Core idea:

1. 💡 The LED emits light at specific wavelengths (commonly green 525 nm or infrared 940 nm).
2. 🌊 The light penetrates the skin and part of it is absorbed by blood.
3. 📉 Changes in blood volume lead to changes in how much light is absorbed.
4. 📊 The detector measures changes in light intensity to form the PPG signal.

Why measure PPG?

PPG signals can provide:

• ❤️ Heart rate: pulse rate (beats per minute)
• 🩸 Blood oxygen saturation (SpO2): combining red and infrared measurements
• 💓 Heart rate variability (HRV): evaluation of autonomic nervous system function
• 🩺 Vascular health: indicators related to arterial stiffness/atherosclerosis
• 💪 Blood pressure estimation: combining other physiological parameters
• 🏃 Exercise intensity: real-time monitoring of exercise status

PPG waveform details

Typical PPG waveform structure

         Systolic peak
           |
           | 
      /----●----\
     /     |     \        Dicrotic notch & peak (diastolic peak)
    /      |      \           /\
   /       |       \         /  \
  /        |        \_______/    \___
  |        |         |      |
  Start    Rise      Decay   Dicrotic notch

Components of the waveform

1. Systolic peak

• Position: the highest point of the waveform
• Physiological meaning: during cardiac contraction, blood rapidly enters the arteries
• Timing: corresponds to the T wave after an ECG
• Characteristics: maximum amplitude and a fast rise rate

2. Dicrotic notch & peak

• Dicrotic notch: a small dip after the systolic peak
• Dicrotic peak: the small peak after the notch
• Physiological meaning: a pressure wave reflection caused by aortic valve closure
• Clinical value: reflects arterial elasticity and peripheral resistance

3. Baseline

• Definition: the lowest point of the waveform
• Meaning: blood volume near end-diastole

Waveform parameters

Parameter	Abbr.	Calculation method	Clinical meaning
Peak-to-peak interval	PP interval	Time interval between two adjacent systolic peaks	Heart rate calculation
Pulse amplitude	PA	Distance from the systolic peak to baseline	Reflects cardiac output and arterial elasticity
Rise time	RT	Time from start to systolic peak	Reflects vascular compliance
Decay time	DT	Time from systolic peak to next start	Reflects peripheral resistance
Dicrotic notch index	DPI	Height of dicrotic peak / height of systolic peak	Vascular aging indicator

Types of PPG signals

Classification by measurement method

1. Transmissive PPG

[LED] → → → [tissue] → → → [detector]
        (light passes through)

Advantages:

• ✅ Strong signal and high SNR
• ✅ Suitable for thin tissues like fingertips and earlobes

Limitations:

• ❌ Not suitable for thick tissues (e.g., wrist)
• 📱 Common applications: medical pulse oximeters, finger clips

2. Reflective PPG

[LED] → → → [tissue]
              ↓ ↓ ↓ (reflected light)
          [detector]

Characteristics:

• ✅ Can be used on any body part
• ✅ Suitable for wearable devices
• ❌ Signal is weaker and more prone to interference
• ⌚ Common applications: smart watches, wristbands, forehead patches

Classification by LED wavelength

Light source	Wavelength	Penetration depth	Main applications
Green light	525 nm	Shallow (1–2 mm)	Heart rate monitoring, HRV analysis
Red light	660 nm	Medium (2–5 mm)	SpO2 measurement (typically paired with infrared)
Infrared	940 nm	Deep (5–10 mm)	SpO2 measurement, deeper vessels

Why do smart watches often use green light?

• 💡 Green light is absorbed the most by blood, making it highly sensitive for heart rate detection
• 🎯 Low energy consumption, suitable for long-term monitoring
• 🌈 Less affected by skin pigment compared with some other wavelengths

Heart rate calculation methods

Method 1: PP interval method (most common)

Heart rate (BPM) = 60 / PP interval (seconds)

Example If PP interval = 0.857 seconds:

Heart rate = 60 / 0.857 ≈ 70 BPM

Method 2: Frequency-domain analysis (FFT)

Steps:
1. Apply FFT transform to the PPG signal
2. Find the main frequency peak in the spectrum
3. Heart rate = main frequency × 60

Advantage: strong noise/interference robustness; suitable for exercise scenarios.

Method 3: Sliding-window counting

Heart rate = (number of peaks in window × 60) / window duration (seconds)

Use case: real-time monitoring with quick response.

Measuring blood oxygen saturation (SpO2)

Principle

SpO2 is estimated using dual-wavelength PPG:

Calculation principle

Oxyhemoglobin (HbO2) and deoxyhemoglobin (Hb) absorb light differently at different wavelengths:

Wavelength	HbO2 absorption	Hb absorption	Characteristics
660 nm (red)	Low	High	Deoxyhemoglobin absorbs more
940 nm (infrared)	High	Low	Oxyhemoglobin absorbs more

SpO2 calculation formula:
SpO2 = A - B × (R/IR)

Where:
- R = AC component / DC component of red light
- IR = AC component / DC component of infrared light
- A, B are empirical coefficients (typically A≈110, B≈25)

Normal range

SpO2 value	Status	Meaning
95–100%	🟢 Normal	Typical range for healthy people
90–94%	🟡 Mild hypoxia	Monitor; may need oxygen
<90%	🔴 Hypoxia	Requires medical intervention
<80%	🚨 Severe hypoxia	Emergency medical situation

PPG signal quality assessment

High-quality signal features

• ✅ Clear and regular waveform

  /\    /\    /\    /\
 /  \  /  \  /  \  /  \
/    \/    \/    \/    \

• ✅ Systolic peak and dicrotic feature are distinguishable
• ✅ Baseline is stable with no large drift
• ✅ Peak-to-peak interval is relatively stable
• ✅ Signal amplitude is moderate (not saturated and not too weak)

Low-quality signal features

• ❌ Motion artifact

   /\  / \/  \
  / \/ \/ \ / \
 /     \/   \/

• ❌ Severe baseline drift
• ❌ Signal is saturated or too weak
• ❌ High-frequency noise interference

Signal quality index (SQI)

# Signal quality evaluation dimensions
SQI = composite_score(
    waveform_completeness,  # 0-1
    signal_to_noise_ratio,  # 0-1
    waveform_regularity,    # 0-1
    amplitude_appropriateness # 0-1
)

# Quality level
if SQI > 0.8:  Excellent
elif SQI > 0.6:  Good
elif SQI > 0.4:  Acceptable
else:  Unusable

Common interferences and solutions

1. Motion artifact

Symptoms:

• Rapid signal fluctuation
• Cannot clearly identify pulse peaks

Causes:

• Body movement
• Relative movement between sensor and skin
• External vibration

Solutions

Hardware:
✅ Improve how the sensor is worn (e.g., adjust wristband tightness)
✅ Use accelerometer to detect motion
✅ Use redundant sensors

Software:
✅ Adaptive filtering (e.g., Kalman filtering)
✅ Independent component analysis (ICA) to separate motion components
✅ Remove motion artifacts based on accelerometer signals
✅ Deep-learning denoising models

2. Poor contact

Symptoms:

• Signal amplitude is too small or zero
• Unstable waveform, intermittent segments

Solutions:

• Ensure the sensor is tight against the skin
• Clean skin surface (remove sweat/oil)
• Adjust the sensor position

3. Ambient light interference

Symptoms:

• Unstable baseline
• Periodic noise superimposed on the signal

Solutions:

• Use a light-blocking cover
• Downsample after high-frequency sampling (>100 Hz)
• Use adaptive ambient light cancellation

4. Temperature effects

Symptoms:

• Signal becomes weaker in cold environments (vessel constriction)
• Baseline drift caused by temperature changes

Solutions:

• Keep the measurement area warm
• Baseline correction algorithms
• Wait a few minutes for the sensor and skin to reach thermal equilibrium

5. Skin pigmentation effects

Symptoms:

• Darker skin may produce smaller signal amplitude

Solutions:

• Increase LED power (follow safety standards)
• Use longer wavelengths (e.g., infrared)
• Use adaptive gain control

Sampling parameter explanation

Selecting sampling rate

Scenario	Recommended sampling rate	Notes
Basic heart rate monitoring	25–50 Hz	Meets basic needs with low power
HRV analysis	100–250 Hz	Needs accurate peak timing
Vascular health assessment	125–500 Hz	Requires fine waveform features
Research applications	500–1000 Hz	High-precision waveform analysis

Important: According to the Nyquist theorem, the sampling rate should be at least 2x the highest frequency component of the signal. While PPG’s main frequencies are often <20 Hz, many systems use >100 Hz to preserve waveform detail.

Signal amplitude

• Unit: arbitrary units (A.U.) or percentage
• Typical range: differs by device; often normalized to 0–100% or 0–1
• Notes:
  • Avoid saturation (too strong causes clipping)
  • Avoid too weak signals (low SNR)

ADC resolution

• 8-bit: 256 levels, for basic applications
• 12-bit: 4096 levels, common in consumer devices
• 16-bit: 65,536 levels, medical-grade accuracy
• 24-bit: 16,777,216 levels, research-grade precision

Advanced applications

1. HRV (Heart Rate Variability) analysis

HRV reflects autonomic nervous system function.

Common HRV metrics

Metric	Meaning	Typical range	Clinical meaning
SDNN	Standard deviation of PP intervals	>50 ms	Overall HRV level
RMSSD	Root mean square of successive PP differences	>30 ms	Parasympathetic activity
pNN50	Proportion of successive differences >50 ms	>10%	Parasympathetic activity
LF	Low-frequency power (0.04–0.15 Hz)	-	Sympathetic + parasympathetic
HF	High-frequency power (0.15–0.4 Hz)	-	Mostly parasympathetic
LF/HF	LF-to-HF ratio	1–3	Balance between sympathetic and parasympathetic

2. Blood pressure estimation

Blood pressure estimation based on PTT (Pulse Transit Time) from PPG:

Blood pressure ∝ 1 / PTT²

PTT = time difference between ECG R-peak and PPG systolic peak

Advantages: non-invasive, continuous monitoring.
Limitations: requires individual calibration; accuracy is limited.

3. Arterial stiffness assessment

Assess vascular elasticity by waveform features from PPG:

# Arterial stiffness indicators
stiffness_index = height / PTT

# Augmentation Index
AIx = (P2 - P1) / PP × 100%

# Where:
# - P1: systolic peak pressure
# - P2: reflected wave pressure
# - PP: pulse pressure (systolic - diastolic)

4. Respiration rate detection

PPG contains respiratory information:

Method 1: baseline changes
Respiration → changes in intrathoracic pressure → changes in venous return → baseline fluctuation

Method 2: amplitude modulation
Respiration affects cardiac output → PPG amplitude varies with the respiratory cycle

Method 3: frequency modulation
Respiratory sinus arrhythmia → PP interval changes with respiration

5. Emotion and stress assessment

Integrate multiple PPG metrics to evaluate psychological state:

Selecting measurement sites

Common sites comparison

Site	Advantages	Disadvantages	Best for
Fingertip	📶 Strong signal ✅ High accuracy	❌ inconvenient for long wear ❌ affected by cold	medical-grade short measurements / diagnosis
Wrist	✅ suitable for long-term wear 💪 convenient daily use	📉 weaker signal 🏃 more motion interference	health monitoring / exercise tracking
Earlobe	📶 relatively strong signal 🎧 doesn’t affect activities	❌ less comfortable ❌ easy to detach	motion monitoring / special use cases
Forehead	✅ vascular-rich 🧠 near the brain	❌ lower social acceptance ❌ affected by sweat	medical monitoring / sleep monitoring
Chest	✅ large vessels ❤️ closer to the heart	❌ needs specialized equipment ❌ not easy for daily wear	professional monitoring / research

Measurement posture recommendations

Resting measurement

• 🪑 Sit or lie down
• 💆 Relax
• 🤚 Keep the measurement site at the same height as your heart
• ⏱️ Measure after 2–5 minutes of rest

During exercise

• 💪 Ensure the sensor is firmly fixed
• ⌚ Use a device that supports exercise mode
• 📊 Check the signal quality indicator after measurement

Recommendations for using our products

Beginners (quick start)

1. Choose the right device

   Recommended: reflective wrist PPG sensor
   - Easy to wear
   - Suitable for daily monitoring
   - Supports heart rate and SpO2

2. Wear it correctly
   • 📏 Adjust wristband tightness (can fit one finger)
   • 📍 Place 2–3 cm behind the wrist bone
   • 🧼 Keep skin clean and dry

Advanced users

1. Use multi-lead (multi-sensor) analysis to evaluate heart function more comprehensively
2. Call advanced API endpoints to obtain detailed analysis results
3. Combine advanced metrics like HRV

Professional institutions

1. Integrate into existing medical systems
2. Use a private deployment solution
3. Follow medical data security standards
4. Perform batch processing and analysis

Important notice

Medical advice

• Our products provide signal analysis and reference suggestions.
• They cannot replace professional medical diagnosis.
• If you observe abnormalities, seek medical care promptly.
• In emergencies, call local emergency services immediately.

Best practices

• Keep measurement conditions consistent (time, posture, etc.)
• Build your personal baseline data for comparisons
• Measure regularly and observe trends
• Save raw data for review and deeper analysis

Related resources

• 📘 PPG signal fundamentals
• 🎓 Best practices for signal collection
• 📊 Signal visualization interpretation guide
• ❓ FAQ

References

1. Wagner GS. Marriott's Practical Electrocardiography (for ECG background; commonly cited in signal processing literature).
2. American Heart Association. ECG Database and Guidelines.