As the advances in imaging modalities in which the image reconstruction problems are mathematically inverse problems, many new kinds of inverse problems have emerged and trends to being high dimensional but with low-cost data acquisition. In order to efficiently and stably solve the under-determined and ill-conditioned inverse problems in high-dimensional medical imaging and compressed sensing, we established accurate statistical models for data fitting and images priors such as shape priors, statistical priors by generative models. In this presentation, I will introduce some of these methods and our recent results for image reconstruction, such as 4DCBCT reconstruction, joint image reconstruction and indirect registration, phase retrieval and some other nonlinear inverse problems.