An Inverted Pinnacle Skyline Query Scheme for Prioritizing Diabetes High-Risk Group Identification
Jun Hyeong Lee, Jongwan Kim · Academic Society for Appropriate Technology · 2025
In this study, we proposed an Inverted Pinnacle Skyline (IPS) query technique utilizing multidimensional glycemic indicators to identify high-risk patients among those with pre-diabetes. Pre-diabetes is a condition where blood glucose levels are higher than normal but do not meet the criteria for diabetes diagnosis, making early intervention and preventive management essential. However, conventional diabetes care systems have primarily focused on treating diagnosed patients, leaving preventive frameworks underdeveloped. To address this gap, the IPS technique was designed to serve as a preventive screening tool by reversing the dominance relationship of the traditional skyline query and identifying skyline points. These skyline points represent patients with glycemic levels closest to the diabetic threshold. We validated the classification performance of the IPS method and demonstrated its effectiveness by comparing it to the traditional linear search method. The IPS technique is expected to contribute as a systematic and efficient tool for preventive diabetes management.