Size-dependent trajectories of DNA macromolecules due to insulative dielectrophoresis in submicrometer-deep fluidic channels.

In this paper, we demonstrate for the first time that insulative dielectrophoresis can induce size-dependent trajectories of DNA macromolecules. We experimentally use lambda (48.5 kbp) and T4GT7 (165.6 kbp) DNA molecules flowing continuously around a sharp corner inside fluidic channels with a depth of 0.4 mum. Numerical simulation of the electrokinetic force distribution inside the channels is in qualitative agreement with our experimentally observed trajectories. We discuss a possible physical mechanism for the DNA polarization and dielectrophoresis inside confining channels, based on the observed dielectrophoresis responses due to different DNA sizes and various electric fields applied between the inlet and the outlet. The proposed physical mechanism indicates that further extensive investigations, both theoretically and experimentally, would be very useful to better elucidate the forces involved at DNA dielectrophoresis. When applied for size-based sorting of DNA molecules, our sorting method offers two major advantages compared to earlier attempts with insulative dielectrophoresis: Its continuous operation allows for high-throughput analysis, and it only requires electric field strengths as low as approximately 10 Vcm.

Spatio-temporal clustering of probabilistic region trajectories

We propose a novel model for the spatio-temporal clustering of trajectories based on motion, which applies to challenging street-view video sequences of pedestrians captured by a mobile camera. A key contribution of our work is the introduction of novel probabilistic region trajectories, motivated by the non-repeatability of segmentation of frames in a video sequence. Hierarchical image segments are obtained by using a state-of-the-art hierarchical segmentation algorithm, and connected from adjacent frames in a directed acyclic graph. The region trajectories and measures of confidence are extracted from this graph using a dynamic programming-based optimisation. Our second main contribution is a Bayesian framework with a twofold goal: to learn the optimal, in a maximum likelihood sense, Random Forests classifier of motion patterns based on video features, and construct a unique graph from region trajectories of different frames, lengths and hierarchical levels. Finally, we demonstrate the use of Isomap for effective spatio-temporal clustering of the region trajectories of pedestrians. We support our claims with experimental results on new and existing challenging video sequences. © 2011 IEEE.