Two-photon calcium imaging is a standard tool for recording large neural populations in vivo, but accurately inferring spikes across the growing diversity of calcium indicators remains an open problem. A new arXiv preprint introduces SpikeSSL, a universal spike inference framework designed to address this challenge.
The framework incorporates dynamics-informed state-space layers. According to the abstract, this design is intended to enable accurate spike inference across different calcium indicators, though the paper's full methodology is not detailed in the abstract excerpt.
The work highlights the need for methods that can keep pace with the expanding range of calcium indicators used in neuroscience research. The preprint is available on arXiv under the identifier 2610.11456.