Projects

Automatic Wireless Health Monitoring System

Heart Rate & Body Temperature Monitoring For Remote Patients

Department of Electrical & Electronic Engineering · Bangladesh University of Engineering & Technology (BUET)

A low-cost, portable wireless patient monitoring system that measures heart rate (BPM) and body temperature (°C) in real time using embedded sensors, RF transmission, and MATLAB signal processing — designed to assist remote patient care in hospital and home settings.

📌 Abstract

Bangladeshis are experiencing heart attacks approximately 10 years earlier than typical sufferers in western countries. Around 40% of all cases occur in people under 50. Constant cardiovascular monitoring is critical yet often unaffordable. This project addresses that gap with a wireless, low-cost device that captures vital signs and transmits them to a doctor or caregiver anywhere in the hospital using RF technology.

⚙️ System Pipeline

[Heart Rate Sensor]  ──┐
                        ├──► [Arduino UNO] ──► [RF Transmitter] ~~wireless~~► [RF Receiver] ──► [Arduino UNO] ──► [PC] ──► [MATLAB GUI]
[DHT11 Temp Sensor]  ──┘

Step-by-step

Sensing — IR pulse sensor and DHT11 temperature sensor continuously read the patient’s vitals.
Processing — Arduino UNO microcontroller reads both sensor outputs and packages the data.
Wireless Transmission — Packaged data is sent via an RF transmitter module.
Reception — A second Arduino connected to the receiver RF module captures the incoming signal.
Post-Processing — Data is fed into MATLAB 2016a where EMD-based signal processing extracts the heart rate.
Display — Final BPM and temperature readings appear on a MATLAB GUI.

🔬 Methodology

Heart Rate Estimation — Empirical Mode Decomposition (EMD)

Raw PPG (photoplethysmography) signals are non-stationary and non-linear. EMD decomposes the signal into a set of Intrinsic Mode Functions (IMFs):

y(n) = Σ(k=1 to N) [ s_k(n) + r_k(n) ]

1000 data points collected at a 500 Hz sampling rate
PPG signal decomposed into IMFs and analysed in the spectral domain
Welch method used for power spectral estimation
IMFs with frequency peaks in the range 0.5–3 Hz are selected and reconstructed
Reconstructed signal is further processed to extract the final heart rate (BPM)

Complete Ensemble EMD (CEEMD) was used over standard EMD to resolve “mode mixing” — the presence of very similar oscillations across different IMF modes.

Temperature Measurement — DHT11 Sensor

The DHT11 communicates with the MCU via a single-wire protocol:

MCU pulls data line LOW for ≥18 ms (Start signal)
MCU pulls HIGH for 20–40 µs, then releases
DHT11 responds: LOW 80 µs → HIGH 80 µs
Sensor transmits 40 bits (5 bytes) of data:

Data (40-bit) = [RH Integer] + [RH Decimal] + [Temp Integer] + [Temp Decimal] + [Checksum]

Checksum = Last 8 bits of (RH_Int + RH_Dec + Temp_Int + Temp_Dec)

Bit encoding:

0 → line HIGH for 26–28 µs after 50 µs LOW
1 → line HIGH for 70 µs after 50 µs LOW

🩺 Sensors

A. Heart Beat Sensor (IR-based, Finger-strap type)

Property	Detail
Type	High-intensity IR reflectance sensor
Principle	Blood-volume pulse changes IR reflectance in capillaries
Form factor	Finger-strap clip
Signal conditioning	Op-amp amplification (very low amplitude raw signal)

B. Temperature Sensor — DHT11

Property	Specification
Supply voltage	3–5.5 V
Temperature range	0–50 °C
Temperature accuracy	±2 °C
Humidity range	20–95% RH
Humidity accuracy	±5% RH
Sampling rate	Max 1 Hz (1 sample/second)
Current draw	2.5 mA max during conversion
Body size	15.5 mm × 12 mm × 5.5 mm
Interface	Single-wire digital (4-pin, 0.1” spacing)

📡 RF Module

Wireless data transmission between the patient-side Arduino and the PC-side Arduino is handled by an RF module pair.

Module	Description
Transmitter	Small PCB subassembly; modulates and transmits data on a radio carrier wave alongside the patient-side Arduino
Receiver	Demodulates the received RF signal; superheterodyne type used for stability across voltage and temperature ranges

Superheterodyne receivers were chosen over super-regenerative for their fixed crystal design, providing better frequency stability and accuracy.

🔧 Hardware — Arduino UNO

Feature	Specification
Microcontroller	ATmega328
Digital I/O pins	14 (6 PWM capable)
Analog inputs	6
Clock speed	16 MHz ceramic resonator
Interface	USB, ICSP header
Power	USB or AC-DC adapter / battery

📊 Results

Signal processing pipeline output:

Raw PPG signal decomposed into 8 IMFs + residual
IMFs in the 0.5–3 Hz cardiac frequency band isolated and reconstructed
Peak-picking algorithm applied to detect heartbeat peaks in the reconstructed signal
Results displayed live in MATLAB GUI

Sample GUI output:

┌──────────────────────────────────────────┐
│         ENTER PATIENT NUMBER:  1         │
│                                          │
│   MEASURE HEART RATE   MEASURE TEMP      │
│                                          │
│        83.28 BPM        38.0 °C          │
│       HEART RATE      TEMPERATURE        │
└──────────────────────────────────────────┘

🌟 Significance

Feature	Benefit
📶 Wireless (RF)	No line-of-sight required; works across hospital wards
💰 Low cost	Affordable for developing-country healthcare settings
🔋 Low power	Minimal battery drain for portable use
📦 Portable	Small form factor; wearable by patients
🏃 Freedom of movement	Patient not tethered to bedside equipment
🧑‍⚕️ Remote monitoring	Doctor can view readings from anywhere in the hospital

🚀 Future Scope

Add ECG and blood pressure monitoring
Integrate pulse oximeter (SpO₂)
Add Galvanic Skin Resistance (stress detection)
GPS integration — automatically notify nearest hospital and dispatch ambulance in emergencies
Auto-call doctor when vitals exceed threshold
Improve RF anti-jamming and data integrity

Supervised by:

Dr. Md. Aynal Haque (Professor, EEE, BUET)
Dr. Mohammed Imamul Hassan Bhuiyan (Professor, EEE, BUET)

Report: Download PDF