Abstract: ‍ ‍Since the 1950s， the research on the electrical activity of neurons has been the focus of many researchers and clinicians. The neurons in the brain form a distributed neural network structure， and different regional structures process different information such as vision， perception， and consciousness activities. And the activities of these neurons are constantly changing over time， thus a method that can comprehensively record the temporal and spatial characteristics of the brain is required. Since its discovery， electroencephalography （EEG） has become one of the most widely used techniques for recording neuronal activities due to its advantages of non-invasiveness， low cost， and high temporal resolution. However， the spatial resolution of EEG is low， which makes it difficult to accurately locate neuronal activities. In order to improve the spatial resolution of EEG， it is necessary to locate the activated neurons of the brain from the measured scalp EEG， that is， to solve the EEG inverse problem. Due to the uncertainty of the electromagnetic inverse solution， the complex activation patterns in the brain may produce the same potential distribution on the scalp. In recent years， studies on the precise anatomical structure and organizational characteristics of the brain and the propagation laws of neuron electrical activities have provided reliable priors for this technology. Numerous studies have demonstrated the significant advantages of this technique in locating brain-sourced activities. Solving the EEG inverse problem can realize the spatial positioning of the electrical activities of neurons through non-invasive means， which is helpful to understand the structure of the distributed neural network and the information transmission processes of the brain. At the same time， it helps to improve the spatial resolution of non-invasive EEG signals and expand the available EEG information dimensions. It has been widely used in neuroscience and technology， clinical medicine and other fields. This paper sorts out the research methods and applications based on the EEG inverse problem. Firstly， it introduces the solution and evaluation methods related to the EEG inverse problem， and then expounds its application progress in the field of brain disease diagnosis and treatment and brain-computer interface. Finally， it summarizes the current problems of this method and analyzes its future development direction.