Three-photon and other higher-order multiphoton imaging modalities are highly sensitive to excitation pulse quality because excitation efficiency scales steeply with instantaneous intensity. The authors address the need to measure pulse dispersion where imaging occurs by developing an in situ approach that estimates dispersion directly at the objective focal plane using third-harmonic generation (THG) interferometric autocorrelation. The technique targets routine pulse characterization across multiphoton microscopy systems without requiring removal of the objective or disruption of the beam path beyond insertion of a compact autocorrelator module.
The reported method records THG interferometric autocorrelation signals obtained at the microscope focal plane. These autocorrelations are interpreted with a Dispersion Look-Up-Table Estimation approach (D-LUTE) combined with a joint two-measurement fitting procedure that extracts baseline dispersion parameters accumulated through the laser and microscope optics. The module is designed to be compact and inserted into the existing beam path of the microscope, enabling measurement at the location of imaging rather than at a remote diagnostic plane.
Prior to experimental application, the authors validated the joint fitting routine using synthetic data. That validation established that the combined D-LUTE and joint two-measurement fitting can reliably recover baseline dispersion parameters such as group-delay dispersion (GDD) and third-order dispersion (TOD) from interferometric autocorrelation traces. The source notes that the fitting procedure was tested on simulated signals to confirm its performance before applying it to physical microscopes.
The method was applied to measurements at excitation wavelengths of 1300 nm and 1600 nm across two microscope systems equipped with different units of the same laser model. At all tested conditions, measured pulse durations at the focal plane exceeded the transform-limit durations by factors of approximately 1.4 to 1.7. The dominant contributor to this broadening was TOD, which the authors found varied by roughly 2.7-fold between the two systems, indicating that dispersion contributions can be strongly system-dependent even when the same laser model is used.
The reported results emphasize that optical components in the laser-microscope path can accumulate GDD and TOD that substantially reduce peak intensity at the focus, and that these quantities are rarely measured in situ without a compact, focal-plane technique.
To demonstrate applicability in vivo, the authors used endogenous THG from myelinated fibers in mouse brain as the signal source for autocorrelation. Using this biological THG, they performed pulse characterization at the tissue surface and at depths up to nearly one millimeter. The in vivo measurements revealed no measurable pulse broadening between the surface and the tested depth range, indicating that, under their conditions, tissue propagation did not further degrade pulse duration detectably over that depth range.
The approach requires only a compact autocorrelator module added to the microscope beam path, making it a low-cost and straightforward option for routine monitoring of pulse dispersion at the focal plane. Because the technique uses THG generated at the focus, it directly probes the relevant excitation conditions for multiphoton microscopy. The source indicates the method is broadly applicable across multiphoton platforms and wavelengths (demonstrated at 1300 nm and 1600 nm) and can be used with endogenous signals for in vivo work.
Limitations explicitly reported in the source are minimal; readers should note this is a preprint and has not undergone peer review. The source reports validation on synthetic data and results from two instrument units; additional validation across more systems and configurations may further define generalizability.
This work is posted as a bioRxiv preprint and has not been certified by peer review. The authors declare no competing interests. The source provides links to supplementary material and data/code; a GitHub repository is referenced for the autocorrelator project. Funding sources listed in the preprint include Howard Hughes Medical Institute, Burroughs Wellcome Fund, and the U.S. National Science Foundation. For details beyond what is summarized here (instrument schematics, fitting code, raw data, parameters), consult the supplementary material and the repository cited by the authors.