Computational modeling and experimental characterization of bacterial microcolonies for rapid detection using light scattering

Nan Bai, Purdue University

Abstract

A label-free and nondestructive optical elastic forward light scattering method has been extended for the analysis of microcolonies for food-borne bacteria detection and identification. To understand the forward light scattering phenomenon, a model based on the scalar diffraction theory has been employed: a bacterial colony is considered as a biological spatial light modulator with amplitude and phase modulation to the incoming light, which continues to propagate to the far-field to form a distinct scattering 'fingerprint'. Numerical implementation via angular spectrum method (ASM) and Fresnel approximation have been carried out through Fast Fourier Transform (FFT) to simulate this optical model. Sampling criteria to achieve unbiased and un-aliased simulation results have been derived and the effects of violating these conditions have been studied. Diffraction patterns predicted by these two methods (ASM and Fresnel) have been compared to show their applicability to different simulation settings. Through the simulation work, the correlation between the colony morphology and its forward scattering pattern has been established to link the number of diffraction rings and the half cone angle with the diameter and the central height of the Gaussian-shaped colonies. In order to experimentally prove the correlation, a colony morphology analyzer has been built and used to characterize the morphology of different bacteria genera and investigate their growth dynamics. The experimental measurements have demonstrated the possibility of differentiating bacteria Salmonella, Listeria, Escherichia in their early growth stage (100∼500 µm) based on their phenotypic characteristics. This conclusion has important implications in microcolony detection, as most bacteria of our interest need much less incubation time (8∼12 hours) to grow into this size range. The original forward light scatterometer has been updated to capture scattering patterns from microcolonies. Experiments have been performed to reveal the time dependent nature of scattering patterns. The experimental work has been compared with simulation results and demonstrated the feasibility of extending this technique for microcolony identification. Lastly, a quantitative phase imaging technique based on the phase gradient driven intensity variation has been studied and implemented to render the 2D phase map of the colony sample.

Degree

Ph.D.

Advisors

King, Purdue University.

Subject Area

Microbiology|Mechanical engineering|Optics

COinS