CAP: Towards PPG Universal Representation Learning with Patient-level Supervision

• URL: http://arxiv.org/abs/2606.15284v1
• Date: Sat, 13 Jun 2026 12:44:27 GMT
• Title: CAP: Towards PPG Universal Representation Learning with Patient-level Supervision
• Abstract summary: Photoplethysmography ( PPG) plays a central role in wearable health monitoring and clinical decision support.<n>Existing approaches to universal PPG representation learning largely focus on signal-level objectives.<n>We propose Clinical Anchored Pretraining for PPG (CAP)<n>CAP performs cross-modal contrastive alignment that anchors PPG representations to patient-level clinical semantics.
• Score: 21.358995073701163
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Photoplethysmography (PPG) plays a central role in wearable health monitoring and clinical decision support. Yet existing approaches to universal PPG representation learning largely focus on signal-level objectives and often overlook patient-level health context, which limits generalization to complex clinical tasks and heterogeneous cohorts. To address this gap, we construct a large-scale paired PPG-EHR multimodal dataset by distilling fragmented medical histories and clinical records into cohesive, patient-level electronic health records (EHR). Building on this resource, we propose Clinical Anchored Pretraining for PPG (CAP). During pretraining, CAP performs cross-modal contrastive alignment that anchors PPG representations to patient-level clinical semantics, guiding the encoder beyond waveform fitting toward modeling consistency in a patient's overall physiological state. During downstream adaptation, the pretrained PPG encoder provides clinically grounded representations that strengthen inductive bias and improve robustness and transferability. Experiments demonstrate that CAP consistently outperforms strong baselines on four diverse downstream tasks. CAP achieves a particularly large gain on respiratory rate prediction (up to +87.6% relative improvement over the state-of-the-art baseline) and delivers an average relative +26.7% across all tasks. We further enhance the interpretability of our approach through comprehensive analyses, including ablations and multiple complementary visualizations of the learned representations. The code for our experiments is available at: https://github.com/gody123gody/CAP .

Related papers

• ProCA: Progressive Contrastive Alignment for Robust EEG Visual Decoding [47.35612571473169]
  EEG visual decoding aims to recover visual semantics from non-invasive neural time-series signals.<n>Existing methods rely on fixed visual or textual anchors whose semantic relations may become misaligned with EEG representations that vary across trials, subjects, and learning stages.<n>We propose Progressive Contrastive Alignment (ProCA), a unified and model-agnostic framework for adaptive neural-semantic alignment.
  arXiv  Detail & Related papers  (2026-09-04T12:48:25Z)
• A robust PPG foundation model using multimodal physiological supervision [7.441186015526675]
  Photoplethysmography is a non-invasive measure of changes in blood volume.<n>Recent PPG foundation models either use open-source ICU datasets with pretraining paradigms that require curated data, or use closed-source field-like PPG data.<n>We propose a PPG foundation model that does not require high-quality or field-like pretraining data, and instead leverages accompanying electrocardiogram and respiratory signals in ICU datasets to select contrastive samples during pretraining.
  arXiv  Detail & Related papers  (2026-06-05T15:08:50Z)
• Towards a General Intelligence and Interface for Wearable Health Data [45.854819784803055]
  We propose a foundation model for wearable health that is pretrained on more than one trillion minutes of unlabeled sensor signals.<n>We find that this population scale representation unlocks label-efficient few-shot learning and generative capabilities for robust daily metric estimation.
  arXiv  Detail & Related papers  (2026-05-21T17:24:06Z)
• Multimodal Graph-based Classification of Esophageal Motility Disorders [73.90451172929117]
  Diagnosing esophageal motility disorders pose significant challenges due to the complexity of high-resolution impedance manometry data and variability in clinical interpretation.<n>This work explores the feasibility of a multimodal Machine Learning-based classification approach that combines HRIM recordings with patient-specific information and incorporates a graph-based modeling of esophageal physiology.
  arXiv  Detail & Related papers  (2026-05-13T14:52:12Z)
• Deriving Health Metrics from the Photoplethysmogram: Benchmarks and Insights from MIMIC-III-Ext-PPG [2.2945816941393766]
  Photoplethysmography (signals) is one of the most widely captured biosignals for clinical prediction tasks.<n>Photoplethysmography (signals) is one of the most widely captured biosignals for clinical prediction tasks.
  arXiv  Detail & Related papers  (2026-03-23T11:19:06Z)
• Reliable Physiological Monitoring on the Wrist Using Generative Deep Learning to Address Poor Skin-Sensor Contact [14.10775902843149]
  Photoplethysmography is a widely adopted, non-invasive technique for monitoring cardiovascular health and physiological parameters in both consumer and clinical settings.<n>While motion artifacts in dynamic environments have been extensively studied, suboptimal skin-sensor contact in sedentary conditions can distort waveform morphology, leading to the loss or misalignment of key features and compromising sensing accuracy.<n>In this work, we propose CP-, a novel framework that transforms Contact Pressure-distorted PPG signals into high-fidelity waveforms with ideal morphology.
  arXiv  Detail & Related papers  (2025-04-03T16:22:15Z)
• Two-Stage Representation Learning for Analyzing Movement Behavior Dynamics in People Living with Dementia [44.39545678576284]
  This study analyzes home activity data from individuals living with dementia by proposing a two-stage, self-supervised learning approach.<n>The first stage converts time-series activities into text sequences encoded by a pre-trained language model.<n>This PageRank vector captures latent state transitions, effectively compressing complex behaviour data into a succinct form.
  arXiv  Detail & Related papers  (2025-02-13T10:57:25Z)
• Evaluating the Fairness of the MIMIC-IV Dataset and a Baseline Algorithm: Application to the ICU Length of Stay Prediction [65.268245109828]
  This paper uses the MIMIC-IV dataset to examine the fairness and bias in an XGBoost binary classification model predicting the ICU length of stay. The research reveals class imbalances in the dataset across demographic attributes and employs data preprocessing and feature extraction. The paper concludes with recommendations for fairness-aware machine learning techniques for mitigating biases and the need for collaborative efforts among healthcare professionals and data scientists.
  arXiv  Detail & Related papers  (2023-12-31T16:01:48Z)
• Adaptive Multi-Agent Deep Reinforcement Learning for Timely Healthcare Interventions [17.405080523382235]
  We propose a novel AI-driven patient monitoring framework using multi-agent deep reinforcement learning (DRL) Our approach deploys multiple learning agents, each dedicated to monitoring a specific physiological feature, such as heart rate, respiration, and temperature. We evaluate the performance of the proposed multi-agent DRL framework using real-world physiological and motion data from two datasets.
  arXiv  Detail & Related papers  (2023-09-20T00:42:08Z)
• Validating polyp and instrument segmentation methods in colonoscopy through Medico 2020 and MedAI 2021 Challenges [58.32937972322058]
  "Medico automatic polyp segmentation (Medico 2020)" and "MedAI: Transparency in Medical Image (MedAI 2021)" competitions. We present a comprehensive summary and analyze each contribution, highlight the strength of the best-performing methods, and discuss the possibility of clinical translations of such methods into the clinic.
  arXiv  Detail & Related papers  (2023-07-30T16:08:45Z)
• Toward Cohort Intelligence: A Universal Cohort Representation Learning Framework for Electronic Health Record Analysis [15.137213823470544]
  We propose a universal COhort Representation lEarning (CORE) framework to augment EHR utilization by leveraging the fine-grained cohort information among patients. CORE is readily applicable to diverse backbone models, serving as a universal plug-in framework to infuse cohort information into healthcare methods for boosted performance.
  arXiv  Detail & Related papers  (2023-04-10T09:12:37Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.