# Maximum Likelihood Expectation Maximization (MLEM) Motivation: Describing one of the [[Radiation Imaging Techniques]] available, for a general case. ## What is Maximum Likelihood Expectation Maximization Maximum Likelihood Expectation Maximization (MLEM) is an iterative statistical process that aims to eventually maximize the [[Expected Value]] in the presence of latent variable, or more basically, an algorithm that maximizes the [[Maximum Likelihood Estimation]]. This process does not assume that the dataset is complete, and as such can be used iteratively to eventually attempt to produce an estimator that yields an estimate as close to the true estimate as possible. There are some general [[Maximum Likelihood Estimator Warnings | warnings]] to consider when using MLEM, which can be found at its respective page. This has seen use in the field of [[Compton Imaging]], and has its own page: [[MLEM for Compton Imaging]]. ## The Two Iterative Steps 1. The Estimation Step - This step estimates the missing variables in the dataset 2. The Maximization Step - This step maximizes the parameters of the model in the presences of the data.