Project
| # | Title | Team Members | TA | Documents | Sponsor |
|---|---|---|---|---|---|
| 44 | PreBP Predict |
Advaith Anand Andrew Cheng Siddharth Salapaka |
Denghan Xiong | ||
| # PreBP Predict Team Members: - Advaith Anand (aanand10) - Andrew Cheng (ac158) - Siddharth Salapaka (svs9) - # Problem Preeclampsia and other hypertensive disorders of pregnancy are usually screened with blood-pressure measurements taken during clinic visits. Because these measurements are only taken at certain times, changes between visits can be missed. PreBP Predict will build a wearable platform for collecting synchronized ECG and PPG measurements over time so that changes in these signals can be studied alongside normal blood-pressure screening. # Solution We propose a small wearable that records ECG and PPG at the same time and sends the data to a Windows PC over Bluetooth Low Energy (BLE). The wearable will handle sensing, sample timing, temporary buffering, and transmission. A Windows application will receive and store the data, display both signals, and perform basic filtering and feature extraction. Building our own hardware gives us access to the raw signals and control over sensor placement and sampling rate. # Solution Components ## ECG Acquisition Subsystem A single-lead ECG circuit will read the heart's electrical signal from skin electrodes and send the waveform to the microcontroller. The ECG front end needs access to the raw waveform, low-noise amplification and filtering, low power use, and an output the microcontroller can read. For the initial design, we plan to use an AD8232ACPZ-R7 ECG front end or a comparable part if availability requires a change. ## PPG Acquisition Subsystem A PPG sensor will use light to measure pulse-related changes in blood volume and provide raw samples that can be lined up with the ECG data. The sensor needs raw sample access, an adjustable sampling rate, low power use, and a digital interface such as I2C or SPI. For the initial design, we plan to use a MAX30102EFD+ PPG sensor or a comparable part if availability requires a change. ## Microcontroller and Synchronization Subsystem A small low-power microcontroller will control the sensors, keep the ECG and PPG samples aligned in time, briefly buffer samples, and prepare the data for BLE transmission. It needs an ADC if required by the ECG front end, I2C or SPI for the sensors, UART or SPI for the BLE module, timers for synchronized sampling, and enough RAM for short buffers. For the initial design, we plan to use an STM32G031F8P6 or a similar small microcontroller in a package that is practical to assemble in the lab. ## BLE Communication Subsystem A separate BLE module will send data from the wearable to the Windows PC. The microcontroller will pass ECG and PPG samples to the module over UART or SPI. The module needs enough data rate for both signals, an integrated BLE stack, and an integrated antenna so the project does not require a custom RF design. For the initial design, we plan to use an RN4871-V/RM118 BLE module or a comparable module. ## Power Subsystem The wearable will run from a small single-cell Li-ion/LiPo battery and provide the required supply voltages to the sensors, microcontroller, and BLE module. The power circuit will include voltage regulation, decoupling, and basic protection, with a low-noise supply for the sensing circuitry. For the initial design, we plan to use an MCP73831T-2ACI/OT charging IC and a TPS63031DSKR 3.3 V regulator, with the battery capacity chosen after measuring the final system current. ## Windows Application The Windows application will connect to the wearable over BLE, receive the ECG and PPG data, save it locally, and display the signals. It will handle BLE scanning and connection, receiving the data through GATT, parsing packets, keeping track of sample timing, and plotting the signals. We will also use it to filter the data, extract useful features, and compare measurements over time. No cloud connection is needed for the prototype. # Criterion For Success The project will be successful if the wearable can record ECG and PPG at the same time for at least 10 minutes, with ECG sampled at at least 200 samples per second and PPG at at least 100 samples per second. After timestamp alignment, the two streams should remain synchronized within 10 ms, and BLE transmission should lose less than 2% of packets during the 10-minute test. The Windows application should correctly receive, store, and display both signals, and the wearable should operate from battery power for at least 2 hours. As a basic signal-quality check, heart rate calculated independently from the ECG and PPG signals should agree within 5 beats per minute during three one-minute resting tests. These tests are intended to verify the sensing platform rather than make a clinical claim about preeclampsia detection. |
|||||