A Multi-level Hypoglycemia Early Alarm System Based on Sequence Pattern Mining
Abstract Background: Early alarm of hypoglycemia, detection of asymptomatic hypoglycemia, and effective control of blood glucose fluctuation make a great contribution to the treatment for diabetic patients. In this study, we designed a multi-level hypoglycemia early alarm system to improve the overall performance of hypoglycemia early alarm. Methods: Through symbolizing the historical CGM data, the Prefix Span was adopted to obtain the early alarm and non-alarm frequent sequence libraries of hypoglycemia. The longest common subsequence was used to remove the common frequent sequence for achieving the hypoglycemia early alarm on the clinical dataset. Frequent sequence pattern libraries with different risk levels were designed by choosing different thresholds and a multi-level hypoglycemia early alarm system for different clinical situations were established. Results: The model was able to predict hypoglycemia events in the clinical dataset of level-I (sensitivity 85.90%, false-positive 23.86%, miss alarm rate 14.10%, average early alarm time 20.61min), level-II (sensitivity 80.36%, false-positive 17.37%, miss alarm rate 19.63%, average early alarm time 27.66min), and level-III (sensitivity 78.07%, false-positive 13.59%, miss alarm rate 21.93%, average early alarm time 33.80min), respectively.Conclusions: The proposed approach could effectively predict hypoglycemia events on the basis of different thresholds to meet different prevention and treatment requirements in clinical situations. Moreover, the experimental results confirm the practicality and prospects of the proposed early alarm system, which reflects further significance in personalized medicine for hypoglycemia prevention.