Abstract
Calcium imaging is becoming an increasingly popular technology to indirectly measure activity patterns in local neuronal networks. Based on the dependence of calcium fluorescence on neuronal spiking, two-photon calcium imaging affords single-cell resolution of neuronal population activity. However, it is still difficult to reconstruct neuronal activity from complex calcium fluorescence traces, particularly for traces contaminated by noise. Here, we describe a robust and efficient neuronal-activity reconstruction method that utilizes Lq minimization and interval screening (IS), which we refer to as LqIS. The simulation results show that LqIS performs satisfactorily in terms of both accuracy and speed of reconstruction. Reconstruction of simulation and experimental data also shows that LqIS has advantages in terms of the recall rate, precision rate, and timing error. Finally, LqIS is demonstrated to effectively reconstruct neuronal burst activity from calcium fluorescence traces recorded from large-size neuronal population.
Highlights
With the development of fluorescent calcium indicators and fast imaging technology, calcium imaging has enabled us to optically detect neuronal activity within local neuronal populations [1,2,3], which is essential for understanding the neural information processing
We proposed the Lq minimization approach and interval screening (LqIS) method for reconstructing the neuronal burst activity from calcium fluorescence traces of a neuronal population
Our reconstruction results show that the LqIS method can suitably reconstruct neuronal activity from calcium fluorescence traces with low signal-to-noise ratio (SNR) and large data size
Summary
With the development of fluorescent calcium indicators and fast imaging technology, calcium imaging has enabled us to optically detect neuronal activity within local neuronal populations [1,2,3], which is essential for understanding the neural information processing. Detecting single spike firing from Ca2+ trace is an easy task because there are no overlapped Ca2+ waves. These non-overlapping Ca2+ waves can be identified with template matching. Reconstruction of burst activity from Ca2+ trace is still a challenge This challenge mainly originates from a large number of noise in Ca2+ imaging of neuronal population [6], high nonlinearities in the spike-calcium relationship [7,8,9], the often relatively low temporal resolution of the calcium signal when compared with the action potential timescale [10]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
