Objective: Statistical models are powerful tools for describing biological phenomena such as neuronal spiking activity. Although these models have been widely used to study spontaneous and stimulated neuronal activity, they have not yet been applied to analyze responses to electrical cortical stimulation. In this study, we present an innovative approach to characterize neuronal responses to electrical stimulation in the mouse cortex, providing detailed insights into cortical-thalamic dynamics. Approach: Our method applies Mixture Models to analyze the Peri-Stimulus Time His togram of each neuron, predicting the probability of spiking at specific latencies following the onset of electrical stimuli. By applying this approach, we investigated neuronal re sponses to cortical stimulation recorded from the motor cortex, somatosensory cortex, and sensorimotor-related thalamic nuclei in the mouse brain. Main results: The characterization approach achieved high goodness of fit, and the model features were leveraged by applying machine learning methods for stimulus intensity decoding and classification of brain regions to which a neuron belongs given its response to the stimulus. The Random Forest model demonstrated the highest F1 scores, achieving 92.86% for stimulus intensity decoding and 84.35% for brain zone classification. Significance: This study presents a novel statistical framework for characterizing neu ronal responses to electrical cortical stimulation, providing quantitative insights into cortical thalamic dynamics. Our approach achieves high accuracy in stimulus decoding and brain region classification, providing valuable contributions for neuroscience research and neuro technology applications