Oxygen comfort evaluation method based on symptom index for short-term internal migrants to Tibet.

In Tibet, the hypobaric-hypoxic environment found at high altitudes leads to dysfunction in short-term internal migrants and has noticeable effects on physiology, psychological health, and comfort level. Therefore, it is essential to accurately determine the degree of hypoxia and improve the hypoxic environment of plateaus. Despite advances in the medical diagnosis and treatment of pathological hypoxic injuries, there are some limitations in the oxygenic evaluation of internal migrants with mild hypoxia. An oxygen comfort evaluation method (OCEM) based on typical anoxic symptomatology and physiological indices is proposed in this study. Experiments with different oxygen concentrations were conducted to measure anoxic symptomatology and physiological indices. Using item and exploratory factor analyses, 19 symptom indices were screened to predict oxygen sensation in humans. Finally, the OCEM was established using an artificial neural network and fuzzy mathematics method and its accuracy was verified through a field survey. The results showed that the artificial neural network model using symptomatologic indices could predict human oxygen sensation, with an area under the receiver operating characteristic curve of 0.630-0.913 and prediction accuracy of 93 %. Oxygen comfort can be predicted from the oxygen sensation and typical physiological indices using the fuzzy mathematics method; the weighted kappa coefficient was 0.825, indicating a strong correlation between the predicted and actual values. The proposed OCEM can help determine the oxygen comfort conditions of high-altitude internal migrants and provide a basis for indoor oxygen environment regulation in high-altitude buildings.