I am currently pursuing my PhD in the Department of Electrical Engineering at Princeton University, United States. My research is to design and fabricate micro-scale devices on silicon wafer capable of sorting blood cells based on their physical properties. The technique is known as Deterministic Lateral Displacement (DLD); I make micro-scale chips with DLD arrays for efficiently separating white blood cells from a very tiny amount of blood (microliters). White blood cells are necessary for a wide variety of research and gene therapies, and usually, white cells are larger in size compared to other cells. My research is oriented to make sure high throughput of white blood cells with low contamination from other cells using my DLD device.

Before starting the project on DLD arrays, I have worked on statistical modeling of Quantum Cascade Laser-based Dual-Comb Spectroscopic (DCS) System at Princeton. The aim of the project was to detect toxic gases in the environment before reaching their harmful levels using spectroscopic absorption data acquired with Dual-comb Spectroscopic system. My work was focused on processing the data, identifying the gases with higher sensitivity and selectivity, and measure the contamination level. The complexity of the system arises due to its developing nature as technology and highly noisy data. I statistically modeled various possible atmospheric and systematic noises available in theory according to my observation from experimental field-data and incorporated those in the statistical model of the DCS system. Then, I used machine learning algorithms to optimize the findings and analyzed the probability of false positives of the system for critical threshold amounts of various gases theoretically for different setups. My theoretical modeling results were able to replicate the experimentally observed data and I developed a tool to analyze the theoretical performance of the DCS system. I also developed an algorithm to determine environmental gases with the least correlated spectroscopic absorption features in a chosen spectroscopic window.

During my stay at the University of Dhaka, I supervised my students on several projects, for example: Handwritten Bengali character recognition using SVM, Human Action Recognition for ubiquitous wearable sensors, and so on.