The study reports on a high-altitude EEG dataset comprising 64-channel EEG signals from 23 subjects, aiming at achieving a finer-grained assessment of hypoxia levels. Four hypoxia levels were induced by creating a gradient of oxygen partial pressure through changes in altitude and external hypoxia stimulation. The dataset was collected in a hypoxic chamber that simulates altitude changes, allowing for a refined classification of different hypoxia levels based on ranges of oxygen saturation. The total recorded EEG data amounts to approximately 10.25 hours. Validation results indicate that the four hypoxia levels can be effectively recognized using EEG signals. Compared to binary classification, our fine-grained dataset allows for more precise detection of hypoxia levels. This dataset is anticipated to have significant research and practical value in developing accurate methods for identifying hypoxia levels. As a valuable and standardized resource, it will enable extensive analysis and comparison for researchers in the field of high-altitude hypoxia.
© 2024. The Author(s).