<p>Electromagnetic radiation (EMR) is a double-edged sword: it may harm living organisms when it exceeds the safe range, while when properly regulated, it can provide critical support for the research and diagnosis of the nervous system. This paper simulates EMR through a memristor model to study the impact of EMR on four-dimensional delayed Hopfield neural networks (HNNs). First, the existence and the dispersion range of infinitely many equilibrium points are presented, and a semi-global exponential stability criterion of the zero equilibrium point is derived through a method based on the system’s solutions. In addition, the differences in the chaotic behavior of the considered systems under different parameter values are analyzed, and it is found that transient chaos is sensitive to time delay, and changes in time delay will destroy the periodic state. What is more, the study finds that adding another EMR on the basis of a fixed one can make the system exhibit multi-periodic phenomena, and by changing the values of different parameters, it is revealed that the increased number of EMR sources and parameter variations not only induce more complex system dynamics, but also suppress chaotic behavior under specific configurations. This implies that the dynamic behaviors of delayed HNNs can be controlled by adjusting the number of EMRs and their parameters. Finally, the correctness of the theory and the practicality of the models under consideration are effectively verified through Multisim and FPGA.</p>