眶额皮层神经元经济价值编码的点过程建模OA

Objective Accurately describing the spiking patterns of neurons in orbitofrontal cortex(OFC)is important for understanding the neural mechanisms underlying economic value encoding and its applications in biomedical engineering.This study aims to establish a point process generalized linear model framework to systematically describe the characteristics of value-related neuronal activity influenced by multidimensional factors and its task specificity,thereby providing theoretical references for neuronal firing regulation technologies.Methods A point process generalized linear model was used to analyze the firing activity of 95 neurons in OFC Area 13 when macaques performed value-related choice tasks and value-free stop-signal tasks.The model simultaneously related each neuron's firing probability to multiple factors,including temporal structure of the trial,movement execution,refractoriness,bursting,and oscillatory dynamics.Cardinal spline functions were employed to describe neuronal task responses and firing history dependence,and differences in neuronal firing patterns between the two tasks were compared.Results In value-free tasks,5.49%of neurons exhibited significant firing history effects without significant temporal distinctions,while the proportion was 4.35%in value-related tasks,suggesting that the long-term history-dependent patterns of OFC neurons are task-specific.As an intrinsic hidden dimension,firing history played an important regulatory role in economic value encoding.Conclusions The established point process generalized linear model can effectively capture the multidimensional features of OFC neuronal firing activity and reveal the important modulatory role of neuronal firing history in economic value encoding.This model provides a systematic framework for characterizing the dynamics of OFC neural activity in economic decision-making and offers methodological reference for understanding the neural mechanisms of economic decisions.