Microfluidics

Real-Time Cell Sorting with FPGA-Accelerated Deep Learning

Real-Time Cell Sorting with Scalable In Situ FPGA-Accelerated Deep Learning

This work presents a teacher-student ML framework using knowledge distillation for label-free lymphocyte classification (T4, T8, B cells) from bright-field images, deployed on FPGA with ultra-low 14.5 µs inference latency - 40× faster than previous state of the art.

MIML: Multiplex Image Machine Learning for High Precision Cell Classification via Mechanical Traits within Microfluidic Systems

This paper presents a novel machine learning framework (MIML) that enhances the precision and specificity of cell classification by integrating label-free cell images with biomechanical data, achieving a 98.3% classification accuracy.

Microphysiologically engineered vessel-tumor model to investigate vascular transport dynamics of immune cells

This paper introduces a 3D vessel-tumor model to investigate immune cell transport dynamics, revealing the role of endothelial glycocalyx in T cell transport and tumor killing during cancer immunotherapy.

Transformer-Based Trajectory Prediction in Microfluidics

Constructed a Transformer-based model for accurately predicting cell trajectories in microfluidic devices, focusing on the optimization of designs such as deterministic lateral displacement (DLD).

Adaptable microfluidic vessel-on-a-chip platform for investigating tumor metastatic transport in bloodstream

This paper identifies biomechanical conditions that enhance circulating tumor cell adhesion, providing insights for the development of anticancer therapies.