Abstract: The importance of radioisotopes in modern nuclear medicine is becoming more and more significant. It is imperative to speed up independent research and development and get rid of the plight of relying on imports of radioisotopes. Nuclide separation is one of the key technologies for radioisotope preparation, and it is very important for engineers to master the separation process to quickly and accurately judge the separation of target nuclides. Aiming at the typical process of radioisotope separation, an online detection software for radioisotope separation is designed and developed based on scintillation detector. The software is used to measure the radioactivity flowing in the pipeline during the separation process, including count rate and radionuclide composition. The measured signal is transmitted to the control system of the separation device for analysis, processing and display. In the nuclide discrimination, the sequential Bayesian nuclide identification algorithm is used, and the Kalman filter is optimized based on the genetic algorithm to achieve better parameter estimation, accurately and quickly identify specific radionuclides in the mixture, and improve the efficiency and efficiency of the nuclide separation process. accuracy. The software realizes the data transmission between the equipment and the system based on the serial port communication, realizes the functions of data acquisition, real-time measurement, visual display, gamma energy spectrum drawing, gamma energy spectrum analysis, system setting, process control, nuclide identification, etc., software interface Simple, experimentally tested, stable operation, friendly human-computer interaction, good practicability and maintainability, providing a technical reference for on-line detection of radioisotope separation and preparation.