2018/03/13 by Yu Terada, Tomoyuki Obuchi, Terada, Yu +5
Neuroscience · #Neural dynamics and brain function #Functional Brain Connectivity Studies #Neuroscience and Neuropharmacology Research
paper · pdf · doi:10.48550/arxiv.1803.04738
Recent remarkable advances in the experimental techniques have provided a\nbackground for inferring neuronal couplings from point process data that\nincludes a great number of neurons. Here, we propose a systematic procedure for\npre- and post-processing generic point process data in an objective manner, to\nhandle data in the framework of a binary simple statistical model, the Ising or\ngeneralized McCulloch--Pitts model. The procedure involves two steps: (1)\ndetermining time-bin size for transforming the point-process data into\ndiscrete-time binary data and (2) screening relevant couplings from the\nestimated couplings. For the first step, we decide the optimal time-bin size by\nintroducing the null hypothesis that all neurons would fire independently, then\nchoosing a time-bin size so that the null hypothesis is rejected with the most\nstrict criterion. The likelihood associated with the null hypothesis is\nanalytically evaluated and used for the rejection process. For the second\npost-processing step, after a certain estimator of coupling is obtained based\non the pre-processed dataset, the estimate is compared with many other\nestimates derived from datasets obtained by randomizing the original dataset in\nthe time direction. We accept the original estimate as relevant only if its\nabsolute value is sufficiently larger than them of randomized datasets. These\nmanipulations suppress false positive couplings induced by statistical noise.\nWe apply this inference procedure to spiking data from synthetic and in vitro\nneuronal networks. The results show that the proposed procedure identifies the\npresence/absence of synaptic couplings fairly well including their signs, for\nthe synthetic and experimental data. In particular, the results support that we\ncan infer the physical connections of underlying systems in favorable\nsituations, even when using the simple statistical model.\n