FPGA-based real-time iterative learning control for bidirectional galvo scanning in large-FOV two-photon microscopy

Two-photon microscopy enables deep, minimally invasive imaging, but high-speed bidirectional scanning is constrained by galvo phase lag and nonlinear dynamics, leading to line-to-line misregistration at large field of view (FOV) that is difficult to correct and severely degrades image quality. Existing model-based correction strategies are computationally intensive and difficult to implement in practice. We introduce a lightweight iterative learning control (ILC) scheme that leverages the periodicity of raster scanning for real-time compensation. Implemented on a field-programmable gate array (FPGA), the algorithm iteratively updates control signals using a proportional ILC law with microsecond latency. This reduces galvo tracking error to ~2% of its open-loop level and completely suppresses odd–even line shifts. Pollen imaging over a 310-μm FOV demonstrates full restoration of structural detail. The proposed FPGA–ILC framework is computationally efficient, easily deployable, and substantially enhances large-FOV two-photon imaging. Beyond two-photon microscopy, it is broadly applicable to other raster-scanning modalities such as confocal microscopy, LiDAR, and laser micromachining.