Benchmarking Open-Source PPG Foundation Models for Biological Age Prediction

• URL: http://arxiv.org/abs/2603.14030v1
• Date: Sat, 14 Mar 2026 17:11:38 GMT
• Title: Benchmarking Open-Source PPG Foundation Models for Biological Age Prediction
• Abstract summary: A task-specific model trained on 212,231 UK Biobank subjects to predict age from PPG (AI- vascular Age) fails on a different clinical population.<n>We investigate why this happens and what it means for PPG-based biological age prediction.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: A task-specific model trained on 212,231 UK Biobank subjects to predict vascular age from PPG (AI-PPG Age) fails on a different clinical population: predictions collapse to a narrow 38-67 year range regardless of true age. Meanwhile, a general-purpose foundation model with no age-related training objective achieves lower error on the same data. We investigate why this happens and what it means for PPG-based biological age prediction. We evaluate three open-source PPG models (Pulse-PPG, PaPaGei-S, AI-PPG Age) on 906 surgical patients from PulseDB, using frozen embeddings with Ridge regression and 5-fold cross-validation. Pulse-PPG reaches MAE = 9.28 years, beating both AI-PPG Age in linear probe mode (9.72) and HR/HRV combined with demographics (9.59). Adding demographic features brings the best result down to MAE = 8.22 years (R2 = 0.517, r = 0.725). The predicted age gap correlates with diastolic blood pressure after adjusting for chronological age (r = -0.188, p = 1.2e-8), consistent with what Apple reported for their proprietary PpgAge model. The remaining gap with Apple (MAE 2.43) appears driven by dataset size (906 vs 213,593 subjects) and population differences rather than model architecture, as our learning curve shows no plateau. Code is publicly available.

Related papers

• STST-JEPA: Shallow-Target Spatio-Temporal Joint Embedding Prediction Architecture For EEG Self-Supervised Learning [0.0]
  Brain age is an emerging biomarker whose deviation from chronological age tracks neurological and psychiatric burden.<n>We introduce STST-JEPA, a self-supervised transformer for resting-state and task EEG, pretrained on 47,703 sessions spanning ages 5-81.
  arXiv  Detail & Related papers  (2026-07-07T12:19:54Z)
• Simulating clinical interventions with a generative multimodal model of human physiology [49.828582089394445]
  HealthFormer is a decoder-only transformer that models the human physiological trajectory generatively.<n>We tokenise each participant's health trajectory across 667 measurements spanning seven domains.<n>HealthFormer improves prediction for 27 of 30 incident-disease and mortality endpoints, exceeding established clinical risk scores in every comparison.
  arXiv  Detail & Related papers  (2026-04-30T14:10:30Z)
• Child Mortality Prediction in Bangladesh: A Decade-Long Validation Study [0.0]
  The Demographic and Health Surveys (DHS) data from Bangladesh for 2011-2022, with n = 33,962, are used in this paper.<n>We trained the model on (2011-2014) data, validated it on 2017 data, and tested it on 2022 data.<n>Eight years after the initial test of the model, a genetic algorithm-based Neural Architecture Search found a single-layer neural architecture to be superior to XGBoost.
  arXiv  Detail & Related papers  (2026-02-03T19:18:50Z)
• Explainable Admission-Level Predictive Modeling for Prolonged Hospital Stay in Elderly Populations: Challenges in Low- and Middle-Income Countries [65.4286079244589]
  Prolonged length of stay (pLoS) is a significant factor associated with the risk of adverse in-hospital events.<n>We develop and explain a predictive model for pLos using admission-level patient and hospital administrative data.
  arXiv  Detail & Related papers  (2026-01-07T23:35:24Z)
• Using Pre-training and Interaction Modeling for ancestry-specific disease prediction in UK Biobank [69.90493129893112]
  Recent genome-wide association studies (GWAS) have uncovered the genetic basis of complex traits, but show an under-representation of non-European descent individuals. Here, we assess whether we can improve disease prediction across diverse ancestries using multiomic data.
  arXiv  Detail & Related papers  (2024-04-26T16:39:50Z)
• How Suboptimal is Training rPPG Models with Videos and Targets from Different Body Sites? [40.527994999118725]
  Most current models are trained on facial videos using contact PPG measurements from the fingertip as targets/ labels. We show that neural models learn to predict the morphology of the ground truth PPG signal better when trained on the forehead.
  arXiv  Detail & Related papers  (2024-03-15T15:20:21Z)
• Community-Based Hierarchical Positive-Unlabeled (PU) Model Fusion for Chronic Disease Prediction [35.76481037888834]
  We present a novel Positive-Unlabeled Learning Tree (PUtree) algorithm. PUtree is designed to take into account communities such as different age or income brackets, in tasks of chronic disease prediction. We demonstrate the superior performance of PUtree as well as its variants on two benchmarks and a new diabetes-prediction dataset.
  arXiv  Detail & Related papers  (2023-09-06T22:16:58Z)
• Prostate Age Gap (PAG): An MRI surrogate marker of aging for prostate cancer detection [0.15518894748362708]
  The Prostate Age Gap (PAG) was significantly associated with the risk of clinically significant PC (csPC) and outperformed other well-established PC risk factors.
  arXiv  Detail & Related papers  (2023-08-10T05:20:25Z)
• Explaining machine learning models for age classification in human gait analysis [10.570744839131775]
  The research question was: Which input features are used by ML models to classify age-related differences in walking patterns? We utilized a subset of the AIST Gait Database 2019 containing five bilateral ground reaction force (GRF) recordings per person during barefoot walking of healthy participants. The mean classification accuracy of 60.1% was clearly higher than the zero-rule baseline of 37.3%. The confusion matrix shows that the CNN distinguished younger and older adults well, but had difficulty modeling the middle-aged adults.
  arXiv  Detail & Related papers  (2022-10-16T13:53:51Z)
• On the explainability of hospitalization prediction on a large COVID-19 patient dataset [45.82374977939355]
  We develop various AI models to predict hospitalization on a large (over 110$k$) cohort of COVID-19 positive-tested US patients. Despite high data unbalance, the models reach average precision 0.96-0.98 (0.75-0.85), recall 0.96-0.98 (0.74-0.85), and $F_score 0.97-0.98 (0.79-0.83) on the non-hospitalized (or hospitalized) class.
  arXiv  Detail & Related papers  (2021-10-28T10:23:38Z)
• An Interpretable Web-based Glioblastoma Multiforme Prognosis Prediction Tool using Random Forest Model [1.1024591739346292]
  We propose predictive models that estimate GBM patients' health status of one-year after treatments. We used total of 467 GBM patients' clinical profile consists of 13 features and two follow-up dates. Our machine learning models suggest that the top three prognostic factors for GBM patient survival were MGMT gene promoter, the extent of resection, and age.
  arXiv  Detail & Related papers  (2021-08-30T07:56:34Z)
• DeepRite: Deep Recurrent Inverse TreatmEnt Weighting for Adjusting Time-varying Confounding in Modern Longitudinal Observational Data [68.29870617697532]
  We propose Deep Recurrent Inverse TreatmEnt weighting (DeepRite) for time-varying confounding in longitudinal data. DeepRite is shown to recover the ground truth from synthetic data, and estimate unbiased treatment effects from real data.
  arXiv  Detail & Related papers  (2020-10-28T15:05:08Z)
• Mini-DDSM: Mammography-based Automatic Age Estimation [0.4932130498861986]
  There is no research done on mammograms for age estimation, as far as we know. Due to lack of public mammography data sets that have the age attribute, we resort to using a web crawler to download thumbnail mammographic images and their age fields from the public data set; the Digital Database for Screening Mammography. Subsequently, we extracted deep learning features from the collected data set, by which we built a model using Random Forests regressor to estimate the age automatically.
  arXiv  Detail & Related papers  (2020-10-01T15:35:11Z)

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.