Key result
An MCNN-ViT model using PPG signals achieves ~91% accuracy for diabetes detection.
Why the study?
To effectively explore the complex nonlinear relationship between PPG signals and diabetes using a feature-fusion automatic detection model.
Does an MCNN-ViT deep learning model using multi-kinetic fusion features of PPG signals accurately detect diabetes in an elderly population?
Case-Control (n=102)
No
Does an MCNN-ViT deep learning model using multi-kinetic fusion features of PPG signals accurately detect diabetes in an elderly population?
A novel deep learning model combining multi-scale convolutional neural networks and vision transformers with PPG signal features demonstrates high accuracy for non-invasive diabetes screening.
May support non-invasive PPG diabetes screening; leaves open prospective validation before clinical adoption.
BACKGROUND: Diabetes is a chronic condition that significantly impacts the cardiovascular system and various other organs. Photoplethysmogram (PPG) signals have been shown to correlate with variations in vascular blood flow and the presence of atherosclerosis. To effectively explore the complex nonlinear relationship between PPG signals and diabetes, we propose an automatic detection model based on the fusion of PPG features. METHODS: The proposed model consists of two main components: 1. Dynamic Fusion Feature Extraction: Short PPG signal window segments are processed using the SGR spatial encoding algorithm to extract dynamic fusion features. 2. Feature Representation Learning: Multi-scale convolutional layers (MCNN) are employed to learn feature representations, while the Vision Transformer (ViT) model is utilized to capture global contextual semantic features. RESULTS: The model was trained and validated on a self-collected medical dataset. The experimental results demonstrate that the classification model, which integrates short time window information, significantly improves detection performance. Specifically, the multi-period sequence input model achieves an accuracy of 91.11%, with a Receiver Operating Characteristic (ROC) curve area of 0.9341, indicating strong diagnostic capability. CONCLUSION: This study is a retrospective case-control study that collected clinical data from three groups of people: those with normal glucose levels, those with poorly controlled diabetes, and those with well-controlled diabetes. The study aims to utilize deep learning algorithms for the early prevention and screening of diabetes.
No takes yet. Share an insight, caveat, or question.
Mingxia Xiao (2025) conducted a case-control in Diabetes (n=102). MCNN-ViT model with multikinetic fusion features of PPG vs. Other deep learning models and single-cycle input was evaluated on Classification accuracy. The MCNN-ViT model combined with multikinetic fusion features of PPG signals achieved an accuracy of 91.11% and an AUC of 0.9341 for diabetes detection.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: