PulseMind: GSR Monitoring System Documentation

Table of Contents

1. System Overview
2. Scientific Background
3. Hardware Implementation
4. Software Architecture
5. Data Processing
6. Visualization System
7. System Flow
8. Mathematical Formulations
9. Mermaid Diagrams
10. References

System Overview

PulseMind is an integrated Galvanic Skin Response (GSR) monitoring system that combines: - ESP32-based hardware sensor - GitHub-based data storage - Web-based visualization dashboard

The system measures electrodermal activity (EDA) as an indicator of sympathetic nervous system arousal, which correlates with psychological stress levels.

Scientific Background

Electrodermal Activity Fundamentals

GSR measures changes in skin conductance (SC) with two components: 1. Skin Conductance Level (SCL): Tonic baseline level (0.05-20 S) 2. Skin Conductance Response (SCR): Phasic responses to stimuli (>0.05 S)

Key physiological relationships: - Sweat gland activity Skin conductance - Sympathetic arousal Sweat production - Stress level SCL magnitude

Based on Boucsein (2012) and Dawson et al. (2011), we classify stress levels:

| Conductance (S) | Stress Level | Physiological State | |------------------|-----------------|------------------------------| | <2.0 | Relaxed | Parasympathetic dominance | | 2.0-5.0 | Normal | Balanced autonomic tone | | 5.0-10.0 | Stressed | Moderate sympathetic arousal | | >10.0 | High Stress | Strong sympathetic activation|

Hardware Implementation

Circuit Design

mermaid graph LR A[Finger Electrodes] --> B[Voltage Divider] B --> C[ESP32 ADC Pin 34] C --> D[3.3V Regulator] D --> E[WiFi Module]

Key parameters: - Electrode voltage: 3.3V DC - Series resistance: 1M - ADC resolution: 12-bit (0-4095) - Sampling rate: 10Hz (100ms interval)

Signal Processing Chain

1. Raw ADC reading (0-4095)
2. Voltage conversion: V = (ADC 3.3) / 4095

3. Conductance calculation: G = (V / R) 10^6 [S] Where R = 1M (constant)

4. Moving average filter (window size=5): G_filtered = (G_i...G_i+4) / 5

Software Architecture

Arduino Code Structure

```mermaid classDiagram class GSRMonitor { +float edaBuffer[5] +String fileSHA +void connectWiFi() +void calibrateGSR() +float readGSR() +String getCurrentSHA() +void uploadData(float conductance) +String getSCLStatus(float conductance) }

class WiFiManager {
    +const char* ssid
    +const char* password
    +void reconnect()
}

class GitHubClient {
    +const char* token
    +const char* repo
    +String getSHA()
    +void updateFile()
}

GSRMonitor --> WiFiManager
GSRMonitor --> GitHubClient

```

Key components: 1. WiFi Connectivity: Manages network connection with automatic reconnection 2. GSR Sensor: Handles calibration and data acquisition 3. GitHub API Client: Manages data storage in repository

Data Flow

```mermaid sequenceDiagram participant Sensor participant ESP32 participant GitHub participant Dashboard

loop Every 30 seconds
    Sensor->>ESP32: Analog Reading
    ESP32->>ESP32: Convert to S
    ESP32->>ESP32: Apply Filter
    ESP32->>GitHub: GET SHA
    GitHub-->>ESP32: Current SHA
    ESP32->>GitHub: PUT New Data
    GitHub->>Dashboard: Data Update
    Dashboard->>Dashboard: Visualize
end

```

Data Processing

Calibration Protocol

1. 30-second baseline measurement (300 samples)
2. Calculate mean baseline conductance: G_baseline = ((V_i)/n) (10^6/R)
3. Adaptive thresholds: G_relaxed = 0.5 G_baseline G_stressed = 2.5 G_baseline

Noise Reduction Techniques

1. Moving average filter: python window = [G1, G2, G3, G4, G5] G_filtered = sum(window) / len(window)
2. Outlier rejection (3 principle): if |G_i - | > 3: discard sample

Visualization System

Dashboard Architecture

```mermaid flowchart TB subgraph Web Dashboard A[Data Fetcher] --> B[Chart Renderer] A --> C[Status Indicator] A --> D[Trend Analyzer] B --> E[GSR Time Series] C --> F[Stress Meter] D --> G[Forecast Model] end

subgraph Data Source
    H[GitHub JSON] --> A
end

```

Key components: 1. Real-time GSR chart (Chart.js) 2. Stress level classification 3. Historical trend analysis 4. Predictive forecasting

Stress Level Calculation

stress_level = if G < 2.0: (G/2.0)25 elif G < 5.0: 25 + ((G-2.0)/3.0)25 elif G < 10.0: 50 + ((G-5.0)/5.0)25 else: 75 + ((G-10.0)/10.0)25

System Flow

Complete System Workflow

mermaid journey title PulseMind System Workflow section Hardware ESP32 Boot: 5: ESP32 WiFi Connect: 3: ESP32 Sensor Calibrate: 8: ESP32 Data Acquisition: 15: ESP32 section Software Data Processing: 10: ESP32 GitHub Sync: 7: Cloud section Visualization Data Fetch: 5: Dashboard Real-time Update: 20: Dashboard User Feedback: 15: User

Mathematical Formulations

Key Equations

1. Conductance Conversion: G(t) = (V(t) 10^6) / R Where:

   • V(t) = Measured voltage at time t
   • R = Fixed 1M resistance

2. Moving Average: G(t) = 1/N G(t-i) for i=0 to N-1 (N=5 in implementation)

3. Stress Score: S(t) = 100 (G(t) - G_min)/(G_max - G_min) Where:

   • G_min = 1.0 S
   • G_max = 20.0 S

4. Recovery Rate: R = (T_recovery / T_total) 100% Where:

   • T_recovery = Time spent below baseline
   • T_total = Total observation time

Mermaid Diagrams

System Architecture

mermaid graph TD A[GSR Sensor] --> B[ESP32] B --> C[WiFi] C --> D[GitHub API] D --> E[Web Dashboard] E --> F[User] F -->|Feedback| A

Data Processing Pipeline

mermaid flowchart LR A[Raw ADC] --> B[Voltage Conversion] B --> C[Conductance Calc] C --> D[Filtering] D --> E[Classification] E --> F[Storage] F --> G[Visualization]

References

1. Boucsein, W. (2012). Electrodermal Activity. Springer Science.
2. Dawson, M. E., et al. (2011). The electrodermal system. Handbook of Psychophysiology.
3. Lykken, D. T., & Venables, P. H. (1971). Direct measurement of skin conductance. Psychophysiology.

This documentation provides a comprehensive technical overview of the PulseMind system, from physiological principles to implementation details. The system demonstrates how physiological signals can be captured, processed, and visualized to provide meaningful stress monitoring.