Two-photon imaging (TPI)
All layers of the pig and human corneas were evaluated. The autofluorescence (AF) intensity revealed morphological aspects of each layer with subcellular resolution. The NAD(P)H AF provides information on the cells’ metabolic activity.
First published report of the comparison between the metabolic activity of corneal epithelial cells, endothelial cells, and keratocytes based on NAD(P)H a1/a2 ratio.
Depth-resolved evaluation of the epithelial layer showing the metabolic changes induced by cell differentiation.
New parameter for retrieving information on the organization of collagen within the stroma, based on FFT analysis of backward detected SHG.
Demonstration that TPI can provide the same morphological information already used to assess the viability of human corneal buttons, with additional information that can aid in sample selection prior to transplantation: it was verified a decrease in the metabolic activity of epithelial and endothelial cells and an increase of keratocytes metabolism with storage time (first report on the alterations induced by storage to the metabolism all corneal cell types)
GLCM and FFT analysis of SHG images can be used to evaluate the organization of collagen within the stroma on corneal buttons. A significant decrease in the organization of the collagen fibers was observed.
Demonstration of the feasibility to diagnose corneal pathologies using TPI through the comparison between healthy corneas and corneas diagnosed with keratoconus, acanthamoeba keratitis, and stromal corneal scars: TPI provides information on the morphological alterations induced by these pathologies, but also on pathology-induced metabolic and structural changes.
One-photon imaging (TPI)
Time-gated fluorescence lifetime microscope operating with 440 nm excitation.
Hi-lo method for sructured illuminatio does not degrade the accuracy of measured lifetimes.
Low corneal FAD signal makes the projection of the sinusoidal patterns required by the structured illumination techniques practically impossible: It was not possible to obtain optical sectioned images of corneal tissues.
Time-Gated microscopy with optical sectioning through structured illumination techniques is not adequate to image FAD in whole corneas. T
RLD provides good accuracy (RME lower than 5.0%) on the free to protein-bound ratio for total counts higher than 10000, if we restrict the gain k to values equal or lower than 5.2. The ratio precision is not degraded by the measurement, being close to the precision defined in the sample.
For low total counts lower than 10000, it is possible to define an optimal region for FAD fluorescence lifetime accuracy and precision: gate separations around 1800 ps and low gains (1.05 or 2.41). The accuracy of the fractional contributions α1 and α2 and of the metabolic ratio α2/α1 remains rather stable for k ≤ 2.41 and Δt >1000 ps, being possible to obtain a ratio accuracy error around 15%. The precision of α1, α2 and of the metabolic ratio almost stabilizes for gate separations between 1500 and 2500 ps, for all k values. In this operating region, the precision of the metabolic ratio is near the limit posed by error propagation.
The accuracy and the precision associated to the value of the metabolic ratio obtained using RLD on a time-gated fluorescence lifetime microscope not only depends on the adopted windowing scheme, but also on the nominal value of the ratio.