Overview: Biophysics of Wrist Photoplethysmography
Photoplethysmography (PPG) is an optical reflectance and transmittance technique used to detect volumetric variations of blood circulating in microvascular tissue beds. By illuminating peripheral tissue with a light-emitting diode (LED) at specific wavelengths (commonly green $\lambda \approx 525\text{ nm}$ or infrared $\lambda \approx 940\text{ nm}$) and measuring backscattered light with a photodiode, sensors construct a pulsatile waveform reflecting vascular hemodynamic parameters.
The composite optical signal consists of two primary components: a constant baseline direct current ($DC$) offset generated by light attenuation through skin tissue, venous blood, non-pulsatile arterial blood, and bone; and an alternating current ($AC$) pulsatile waveform driven by cardiac cycle ejection. The signal-to-noise ratio ($SNR$) quantifies the amplitude integrity of this pulsatile $AC$ signal against thermal noise, ambient lighting interference, microvascular scattering, and motion artifacts.
This simulation directly operationalizes empirical coefficients from Charlton et al. (2025) (*PLOS Digital Health*, e0000585), quantifying wrist photoplethysmography performance across demographic and optical variables. Signal-to-Noise Ratio ($SNR$) is calculated in decibels ($\text{dB}$) using the statistical multivariable regression framework:
$$SNR(\text{dB}) = \text{Base}_{SNR} + \beta_{\text{age}} \cdot (\text{Age} - 20) + \beta_{\text{skin}} \cdot (\text{Fitzpatrick} - 1) + \Delta_{\text{LED}} - \eta_{\text{motion}}$$
Where $\text{Base}_{SNR} = 16.0\text{ dB}$ represents baseline performance at age 20 on Fitzpatrick Type I skin, $\beta_{\text{age}} = +0.12\text{ dB/year}$ models age-related arterial wall stiffening and amplitude augmentation, $\beta_{\text{skin}} = -0.52\text{ dB/level}$ captures epidermal melanin light absorption, $\Delta_{\text{LED}}$ represents compensatory optical power scaling up to $+5.0\text{ dB}$, and $\eta_{\text{motion}}$ represents kinetic artifact interference.
Technical Details: Mathematical Modeling & Architecture
The interactive visualizer employs a decoupled rendering loop executing on the HTML5 Canvas API normalized against the physical device pixel ratio ($DPR = \text{window.devicePixelRatio}$). The dual-view display presents an anatomical cross-sectional wrist optical propagation model alongside a real-time oscilloscope wave recorder.
Synthetic PPG waveforms are synthesized using a superposition of primary cardiac ejection sinusoids and secondary dicrotic wave reflections:
$$S(t) = A_{AC} \cdot \left[ -\cos(2\pi f t) + \alpha_{\text{notch}} \cdot \sin^{15}(\pi f t) \right] + W(t) + N(t)$$
Where $f = 1.2\text{ Hz}$ represents heart rate ($72\text{ BPM}$), $A_{AC} = 5 + 45 \cdot \text{Normalized}(SNR)$ scales pulse amplitude, $\alpha_{\text{notch}} = 0.5 \cdot (1 - \frac{\text{Age}-20}{65})$ modulates dicrotic notch dampening across age bands, $W(t) = \text{Wander} \cdot \sin(0.3 t)$ models low-frequency respiratory baseline oscillation, and $N(t)$ represents zero-mean Gaussian white noise inversely proportional to calculated $SNR$.
The Web Audio API synthesis engine operates asynchronously. Tone frequencies are updated via an exponential ramp bounded between $200\text{ Hz}$ (low SNR warning) and $1000\text{ Hz}$ (high SNR fidelity). State management isolates baseline user configurations using explicit memory snapshots, guaranteeing zero-state corruption during interactive interrupt breaking in Demo mode.
Future Directions & Clinical Applications
Future extensions of this interactive framework will expand physiological modeling parameters to include multi-wavelength pulse oximetry ($R = \frac{(AC/DC)_{\text{red}}}{(AC/DC)_{\text{IR}}}$) for accurate continuous functional oxygen saturation ($\text{SpO}_2$) synthesis. Integrating continuous blood pressure estimation via Pulse Arrival Time ($\text{PAT}$) and Pulse Transit Time ($\text{PTT}$) will enable exploration of non-invasive vascular age monitoring.
Additionally, incorporating machine learning noise-reduction architectures—such as adaptive LMS filters, deep spatial autoencoders, and independent component analysis (ICA)—will allow users to test signal reconstruction algorithms against extreme motion degradation in real time.