Research & Projects

AI-Generated Picture

Project 1

A Smartphone-Based Device for Cuff-Less and Calibration-Free Blood Pressure Monitoring

This project, funded by the National Institutes of Health, aims to develop a smartphone-based device capable of measuring blood pressure without the need for cuffs or calibration. Collaborating with a team of researchers and engineers, I designed algorithms in MATLAB and Python to process PPG waveforms, allowing accurate blood pressure extraction through deep learning. My role involved developing a finger squeezer device with a user-friendly interface, implementing a PID controller for precise control, and optimizing the PPG and ECG system circuitry.

Objectives

The objective was to create a calibration-free blood pressure monitoring solution that could integrate with smartphones. We sought to provide an accessible tool for continuous health monitoring. Key considerations included achieving high accuracy, maintaining user comfort, and ensuring system reliability.

Results

The project successfully produced a functional prototype capable of accurate blood pressure measurements. Notable outcomes included insights into the challenges of calibration-free devices and the effectiveness of deep learning in extracting vital health metrics. The device holds potential for widespread health applications, and the methodology developed here could impact future wearable health technology.

AI-Generated Picture

Project 2

Smart Cuff: Multi-Parameter Hemodynamic Monitoring via a Single Convenient Device

As the primary study coordinator for this NIH-funded project, I led the development of a multi-functional cuff blood pressure monitor. This project involved designing a cuff that could monitor multiple hemodynamic parameters, employing commercial components and custom PCB designs. The project was a collaboration with researchers and clinicians aimed at advancing BP measurement technology to improve patient outcomes.

Objectives

The goal was to design a versatile BP monitor capable of precise, multi-parameter hemodynamic measurements. Critical considerations included maintaining accuracy, integrating mathematical models (such as Dr. Drzewiecki's static model), and ensuring the device could self-calibrate through genetic algorithms in Python.

Results

We developed a multi-functional, self-calibrating BP monitor. This project provided new insights into BP measurement challenges and established a baseline for future device improvements. The prototype's success suggests a promising path for broader applications in clinical and home-based health monitoring.

AI-Generated Picture

Project 3

Non-Invasive Detection of Fluid Overload and Fluid Status in Healthy Participants

This NIH-sponsored project focused on non-invasive methods for monitoring fluid status, particularly in postpartum women. Collaborating with UPMC’s Magee-Womens Hospital, I gathered physiological data to support fluid status assessment and used signal processing and machine learning techniques to interpret biomarkers like BNP.

Objectives

The objective was to assess and monitor fluid levels in postpartum patients non-invasively, aiming to reduce complications from fluid overload. Key considerations included data variability and the need for highly accurate classification based on biomarker analysis.

Results

This project demonstrated the potential of non-invasive fluid status monitoring and highlighted the challenges associated with interpreting physiological biomarkers. The insights gained informed further refinement of our approach, advancing non-invasive diagnostics for clinical settings.

AI-Generated Picture

Project 4

Signal Processing for Automated Driving Systems

As a project manager at Tosan (Parsian) Company, I led a team to develop radar signal processing technology for automated driving systems. My responsibilities included FPGA programming, configuring RADAR ICs, and designing a 14-layer PCB for signal processing. The project involved creating a system capable of accurate distance and object detection for vehicles.

Objectives

Our objective was to develop reliable radar signal processing for automated driving applications. This required creating a robust signal processing pipeline and ensuring system integration with various automotive sensors.

Results

We delivered a functional radar processing system with advanced detection capabilities. The project provided essential experience in radar IC configuration and data analysis through machine learning. The system's success has laid the groundwork for further research into autonomous driving solutions.

Page updated

Google Sites

Report abuse