Automated multi-parametric sorting of micron-sized particles via multi-trap laser tweezers
Kaputa, Daniel S.
MetadataShow full item record
The capabilities of laser tweezers have rapidly expanded since the first demonstration by Ashkin and co-workers in 1970 of the ability to trap particles using optical energy. Laser tweezers have been used to measure piconewton forces in many biological and material science application, sort bacteria, measure DNA bond strength, and even perform microsurgery. The laser tweezers system developed for this dissertation foreshadows the next generation of laser tweezer systems that provide automated particle sorted based upon multiple criteria. Many laser tweezer sorting applications today entail the operator sorting cells from a bulk sample, one by one. This dissertation demonstrates the technologies of pattern recognition and image processing that allow for an entire microscope slide to be sorted without any operator intervention. We already live in an automated world where the cars we drive are built by machines instead of humans. The technology is there, and the only factors limiting the advancements of fully automated biological instrumentation is the lack of developers with the appropriate knowledge sets. This dissertation introduces the concept of sorting particles via a multi-parametric approach where several parameters such as size, fluorescence, and Raman spectra are used as sorting criteria. Since the advent of laser tweezers, several groups have demonstrated the ability to sort cells and other particle by size, or by fluorescence, or by any other parameter, but to our knowledge there does not exist a laser tweezer sorting system that can sort particles based upon multiple parameters. Sorting via a single parameter can be a severe limitation as the method lacks the robustness and class specificity that exists when sorting based upon multiple parameters. Simply put, it makes more sense to determine the worth of a baseball card by considering it's condition as well as it's age, rather then solely upon its condition. By adding another parameter such as the name of the player in the card, one can start collecting Babe Ruth rookie cards instead of mint condition cards of bench warmers. In the future, even better multi-parametric laser tweezer particle sorting systems will be developed that make use of pulsed radiation in order to stimulate nonlinear optical phenomena. This dissertation discusses the feasibility of combining a rapid, non-invasive chemical imaging technology called coherent anti-Stokes Raman scattering (CARS) with a laser tweezer sorting system. This would allow for the birth of a laser tweezer particle sorting system of unprecedented speed and chemical specificity the likes of which the world has not yet seen.