Adaptive optics microspectrometer for cross-correlation measurement of microfluidic flows.

Mapping flows in vivo is essential for the investigation of cardiovascular pathologies in animal models. The limitation of optical-based methods, such as space-time cross correlation, is the scattering of light by the connective and fat components and the direct wave front distortion by large inhomogeneities in the tissue. Nonlinear excitation of the sample fluorescence helps us by reducing light scattering in excitation. However, there is still a limitation on the signal-background due to the wave front distortion. We develop a diffractive optical microscope based on a single spatial light modulator (SLM) with no movable parts. We combine the correction of wave front distortions to the cross-correlation analysis of the flow dynamics. We use the SLM to shine arbitrary patterns of spots on the sample, to correct their optical aberrations, to shift the aberration corrected spot array on the sample for the collection of fluorescence images, and to measure flow velocities from the cross-correlation functions computed between couples of spots. The setup and the algorithms are tested on various microfluidic devices. By applying the adaptive optics correction algorithm, it is possible to increase up to 5 times the signal-to-background ratio and to reduce approximately of the same ratio the uncertainty of the flow speed measurement. By working on grids of spots, we can correct different aberrations in different portions of the field of view, a feature that allows for anisoplanatic aberrations correction. Finally, being more efficient in the excitation, we increase the accuracy of the speed measurement by employing a larger number of spots in the grid despite the fact that the two-photon excitation efficiency scales as the fourth power of this number: we achieve a twofold decrease of the uncertainty and a threefold increase of the accuracy in the evaluation of the flow speed.

Spatiotemporal Image Correlation Analysis for 3D Flow Field Mapping in Microfluidic Devices.

Microfluidic devices reproducing 3D networks are particularly valuable for nanomedicine applications such as tissue engineering and active cell sorting. There is however a gap in the possibility to measure how the flow evolves in such 3D structures. We show here that it is possible to map 3D flows in complex microchannel networks by combining wide field illumination to image correlation approaches. For this purpose, we have derived the spatiotemporal image correlation analysis of time stacks of single-plane illumination microscopy images. From the detailed analytical and numerical analysis of the resulting model, we developed a fitting method that allows us to measure, besides the in-plane velocity, the out-of-plane velocity component down to vz ≅ 65 µm/s. We have applied this method successfully to the 3D reconstruction of flows in microchannel networks with planar and 3D ramifications. These different network architectures have been realized by exploiting the great prototyping ability of a 3D printer, whose precision can reach few tens of micrometers, coupled to poly dimethyl-siloxane soft-printing lithography.

Spatiotemporal image correlation analysis of blood flow in branched vessel networks of zebrafish embryos.

Ramification of blood circulation is relevant in a number of physiological and pathological conditions. The oxygen exchange occurs largely in the capillary bed, and the cancer progression is closely linked to the angiogenesis around the tumor mass. Optical microscopy has made impressive improvements in in vivo imaging and dynamic studies based on correlation analysis of time stacks of images. Here, we develop and test advanced methods that allow mapping the flow fields in branched vessel networks at the resolution of 10 to 20 µm. The methods, based on the application of spatiotemporal image correlation spectroscopy and its extension to cross-correlation analysis, are applied here to the case of early stage embryos of zebrafish.

Image Cross-Correlation Analysis of Time Varying Flows.

In vivo studies of blood circulation pathologies have great medical relevance and need methods for the characterization of time varying flows at high spatial and time resolution in small animal models. We test here the efficacy of the combination of image correlation techniques and single plane illumination microscopy (SPIM) in characterizing time varying flows in vitro and in vivo. As indicated by numerical simulations and by in vitro experiments on straight capillaries, the complex analytical form of the cross-correlation function for SPIM detection can be simplified, in conditions of interest for hemodynamics, to a superposition of Gaussian components, easily amenable to the analysis of variable flows. The possibility to select a wide field of view with a good spatial resolution along the collection optical axis and to compute the cross-correlation between regions of interest at varying distances on a single time stack of images allows one to single out periodic flow components from spurious peaks on the cross-correlation functions and to infer the duration of each flow component. We apply this cross-correlation analysis to the blood flow in Zebrafish embryos at 4 days after fertilization, measuring the average speed and the duration of the systolic and diastolic phases.