A compact optical fibre trap based on a GRIN microlens and a principal component analysis algorithm for trap characterization in noisy environments (Conference Presentation)

Abstract

Widespread use of optical manipulation in combination with advanced imaging techniques will be accelerated by compact, optically simple approaches which are readily integrated into advanced microscopy platforms. For example, optical manipulation has been combined with confocal, multi-photon, and STED microscopes. However, these typically require addition of optical components into the existing beam paths of the microscope, increasing complexity and potentially compromising image quality. Optical fiber trapping (OFT) offers an ultra-compact and simple solution but compromises on trap quality due to the low numerical aperture (NA) and short manipulation distance of optical fibers. Tapered fibers can be fabricated but this further reduces the manipulation distance and requires access to specialist fabrication facilities. Here we present a compact, single-beam, high NA OFT probe design based on a graded-index (GRIN) micro-objective lens and single-mode fiber. The OFT probe uses only off-the-shelf components, enables optical trapping at a distance of 200μm from the probe tip, and is compatible with inverted imaging systems. A challenge with specialist imaging systems is the incompatibility between the specialist imaging modality of the platform and the imaging modality required for trap characterization, resulting in noisy and poor trap characterisation data. To overcome this challenge, we developed an adaptive image filter based on principal component analysis (PCA). The filter separates orthogonal degrees of motion in trap characterisation movies and strong stochastic noise can be removed before tracking, resulting in accurate characterisation. We demonstrate the use of this PCA image filter for in situ characterisation of the GRIN lens OFT probe.