High atomic number element nanoparticles have shown potential in tumor diagnosis and therapy. X-ray fluorescence computed tomography (XFCT) technology enables quantitative imaging of high atomic number elements by specifically detecting characteristic X-ray signals. The potential for further biomedical applications of XFCT depends on balancing sensitivity, spatial resolution, and imaging speed in existing XFCT imaging systems. In this study, we utilized a high-energy resolution pixelated photon-counting detector for XFCT imaging. We tackled degradation caused by multi-pixel events in the photon-counting detector through energy and interaction position corrections. Sensitivity and spatial resolution imaging experiments were conducted using PMMA phantoms to validate the effectiveness of the multi-pixel events correction algorithm. After correction, the system's sensitivity and spatial resolution have both improved. Furthermore, XFCT/CBCT dual-modality imaging of gadolinium nanoparticles within mice subcutaneous tumor was successfully achieved. These results demonstrate the preclinical research application potential of the XFCT/CBCT dual-modality imaging system in high atomic number nanoparticle-based tumor diagnosis and therapy.