Abstract: Tumor tissues exhibit pronounced spatial heterogeneity and a complex microenvironment, making it difficult for traditional bulk-level analytical approaches such as transcriptomics and proteomics to capture the intricate cellular interactions and spatial distribution patterns within tumors. In recent years, spatial proteomics methods with spatially resolved capabilities have emerged. Nontargeted techniques allow for panoramic molecular detection but are often limited by insufficient sensitivity and specificity, while conventional targeted approaches provide multiplexing capacity and specificity but remain constrained by throughput, quantitative accuracy, and spatial resolution. To overcome these limitations, mass spectrometry imaging (MSI) has been developed, with representative platforms including imaging mass cytometry (IMC), multiplex ion beam imaging (MIBI), matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI), and secondary ion mass spectrometry (SIMS) imaging. These platforms offer unique advantages by achieving subcellular-level resolution, enabling simultaneous multi-target quantification, and delivering robust quantitative performance, thereby addressing key challenges in spatial proteomics research. The data acquisition and analysis workflow of MSI typically involves standardized preparation of tumor samples, target-specific labeling and multiplex staining, point-by-point imaging through laser or ion beams, signal preprocessing, machine learning or deep learning-based cell segmentation and phenotype annotation, and higher-order spatial structure analyses. Through these processes, antibody-based MSI enables the detailed mapping of cellular architecture and interactions within tumor tissues. Clinical studies have demonstrated its ability to uncover spatial interaction networks among immune cells, tumor cells, and stromal cells, as well as to identify spatial microdomains associated with prognosis and treatment response. These findings not only contribute to the discovery of tumor spatial biomarkers but also provide valuable evidence for treatment evaluation, ultimately facilitating the optimization of precision oncology strategies. Looking ahead, further advancements will require the optimization of high-throughput data acquisition workflows, integration of multi-platform spatial omics data, incorporation of dynamic spatiotemporal imaging techniques, and the development of unified artificial intelligence-driven analytical frameworks. Together, these innovations will enable multimodal characterization of the tumor microenvironment and accelerate the translation of MSI into clinical applications, thereby advancing personalized cancer diagnosis and therapeutics.