In recent years there has been explosive growth in imaging studies in medical and public health research. It is of great interest to understand how subject-level characteristics, including clinic variables and genetic factors, influence imaging phenotypes. In this talk, the speaker will present a class of image-on-scalar regression models for imaging responses and scalar predictors. Through flexible multivariate splines over triangulations, the proposed method can handle the irregular domain of the objects of interest on the images and other characteristics of images. The proposed method can provide the pointwise confidence intervals and data-driven simultaneous confidence corridors to access the estimation uncertainty and automatically identify significant regions. The proposed method is applied to the spatially normalized Positron Emission Tomography data of Alzheimer's Disease Neuroimaging Initiative.