PPG models get better at wearable health data with pruning and fine-tuning
Unlocking the Specialization Capability of PPG Foundation Models for Wearable Signals through Structured Pruning and Adaptive Fine-Tuning
Artificial IntelligenceMachine Learning
Summary
Wearable health devices collect signals called PPG to monitor heart and blood flow. Current computer models trained mostly on hospital data struggle to work well with noisier data from wearables. The authors found that these models have extra parts they don't always use, so they created a way to trim the unneeded parts before fine-tuning the model. This makes the models better suited for different health tasks using wearable data, often beating previous methods. Their approach could help build smarter health monitors that work well in everyday settings.
What this means in practice
- •For wearable device engineers: Improve health monitoring by customizing PPG models to work better with noisy, real-world sensor data.
- •For biomedical signal analysts: Apply structured pruning before fine-tuning to enhance prediction accuracy on diverse health datasets from wearables.
Authors
Huaming Du, Zhixian Xie, Gaoyang Zhen, Gang Kou, Guisong Liu, Yanchao Tan, Wendan Tao, Yuhao Xu, Carl Yang
Abstract
Photoplethysmography (PPG) plays a central role in wearable health monitoring and clinical decision support. However, existing approaches to universal PPG representation learning are predominantly trained on high-quality large-scale datasets obtained from oximeters in hospital settings and often struggle to adapt to sparse and low-quality real-world datasets, thereby limiting their generalization to clinically relevant tasks on prevalent PPG data such as collected from wearable devices. Surprisingly, even after full fine-tuning, PPG foundation models cannot consistently outperform general-purpose time-series models. Therefore, a key question is how to effectively adapt PPG foundation models to target prediction tasks. Through empirical studies on various PPG foundation models, we find that pretrained models often exhibit inherent sparsity and computational redundancy, suggesting that PPG foundation models have learned to activate task-relevant network substructures for different prediction tasks. To preserve this valuable prior knowledge, we propose a structured pruning-and-fine-tuning framework, SPAFT, which regularizes the adaptive fine-tuning process by constraining it to a more relevant and compact parameter space based on parameter importance and structural redundancy. Extensive experiments across four PPG foundation models, five benchmark datasets, and ten downstream tasks demonstrate that SPAFT achieves win rates of 82.8% and 87.5% on regression and classification tasks, respectively. This prune-then-finetune paradigm often enables PPG foundation models to achieve state-of-the-art performance, laying a promising foundation for the real-world clinical deployment of PPG foundation models and the development of personalized PPG foundation models.