Supervised learning sets benchmark for robust spike detection from calcium imaging signals
This addresses a fundamental challenge in calcium imaging for neuroscientists, providing a robust benchmark and improved algorithm for spike detection.
The authors tackled the problem of inferring action potential timing from noisy calcium fluorescence traces by systematically evaluating spike inference algorithms on a large benchmark dataset from neural tissue. They showed that a new supervised learning algorithm outperforms all previously published techniques, setting a new standard with better performance even on unseen datasets.
A fundamental challenge in calcium imaging has been to infer the timing of action potentials from the measured noisy calcium fluorescence traces. We systematically evaluate a range of spike inference algorithms on a large benchmark dataset recorded from varying neural tissue (V1 and retina) using different calcium indicators (OGB-1 and GCamp6). We show that a new algorithm based on supervised learning in flexible probabilistic models outperforms all previously published techniques, setting a new standard for spike inference from calcium signals. Importantly, it performs better than other algorithms even on datasets not seen during training. Future data acquired in new experimental conditions can easily be used to further improve its spike prediction accuracy and generalization performance. Finally, we show that comparing algorithms on artificial data is not informative about performance on real population imaging data, suggesting that a benchmark dataset may greatly facilitate future algorithmic developments.