Mesoscope imaging enables the recording of neural activity projections in the dorsal cortex of behaving subjects through photon excitation and fluorescent indicators that measure intracellular calcium movements. Here, we introduce comprehensive methods for analyzing mesoscope imaging data. This chapter begins with the essential preprocessing steps, including normalization, denoising, and hemodynamic correction. Various decomposition techniques to characterize the spatial and temporal information of mesoscope imaging data are covered, such as region of interest analysis, independent component analysis, singular vector decomposition, nonnegative matrix factorization, and localized semi-nonnegative matrix factorization. Additionally, this chapter explores spatiotemporal flow analysis methods to further understand the dynamics within mesoscope imaging data. The functional connectivity derived from mesoscope imaging data reveals the interaction between different brain regions. This chapter introduces key measurements of functional connectivity, including correlation, centrality, and Granger causality. Artificial intelligence models trained on neural activity and behavior can uncover the hidden connections between neural activity and behavior, while offering profound insights into cognitive and motor-related tasks. This chapter highlights effective behavior decoding models such as linear regression, support vector machines, and recurrent neural networks. Moreover, we discuss neural encoding studies that reveal how sensory information is translated into neural activity patterns in the brain. We end with open questions in the analysis of mesoscopic imaging data.