PACD-Net: Pseudo-Augmented Contrastive Distillation for Glycemic Control Estimation from SMBG
PACD-Net is a new framework designed to improve glycemic control estimation from self-monitoring of blood glucose (SMBG) data. It utilizes a self-supervised contrastive knowledge distillation approach to enhance the accuracy of metrics like Time in Range. The model demonstrates superior performance compared to existing methods, particularly in scenarios with sparse data.
- ▪Effective diabetes management requires continuous monitoring of glycemic levels.
- ▪PACD-Net uses pseudo-SMBG samples to guide learning from sparse observations.
- ▪The model outperforms existing methods in estimating glycemic control metrics from real-world SMBG data.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Machine Learning arXiv:2605.20751 (cs) [Submitted on 20 May 2026] Title:PACD-Net: Pseudo-Augmented Contrastive Distillation for Glycemic Control Estimation from SMBG Authors:Canyu Lei, David Repaske, Jianxin Xie View a PDF of the paper titled PACD-Net: Pseudo-Augmented Contrastive Distillation for Glycemic Control Estimation from SMBG, by Canyu Lei and 2 other authors View PDF HTML (experimental) Abstract:Effective diabetes management requires continuous monitoring of glycemic levels. Clinically, glycemic control is assessed using metrics such as Time in Range (TIR), Time Below Range (TBR), and Time Above Range (TAR), typically derived from continuous glucose monitoring (CGM).
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
