{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,7]],"date-time":"2026-08-07T16:52:27Z","timestamp":1786121547187,"version":"build-2736575974"},"reference-count":52,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,7]],"date-time":"2026-08-07T00:00:00Z","timestamp":1786060800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Pain assessment in non-communicative populations\u2013particularly neonates and cognitively impaired patients\u2013remains a critical clinical challenge, as current automated methods require labeled pain datasets that are both ethically problematic and scarce for vulnerable populations.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We propose a framework trained on zero labeled real pain examples from the target population, combining synthetic pain simulation with unsupervised domain adaptation. Using latent diffusion models, we generate 50,000 realistic pain scenarios spanning facial expressions, physiological signals (heart rate variability and electrodermal activity), and temporal dynamics across diverse demographics. A transformer-based architecture integrates these multimodal cues, and a three-stage domain alignment procedure (contrastive learning, adversarial adaptation, and consistency regularization) bridges the synthetic-to-real gap using only unlabeled real data. We explicitly distinguish this synthetic-source unsupervised domain adaptation setting from classical zero-shot learning (Section 2.4).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      Evaluation across three benchmarks shows competitive performance: MAE of 0.89 on the UNBC-McMaster dataset (31% gap vs. supervised methods), 78.3% accuracy on the BioVid Heat Pain Database (10.5% gap), and a preliminary point estimate of 81.5% accuracy for neonatal assessment (12.7 percentage points above a clinical AU-based baseline, with wide uncertainty given the limited\n                      <jats:italic>N<\/jats:italic>\n                      = 120 cohort; prospective validation on a larger cohort is planned). The proposed approach achieves 40.4% better cross-dataset generalization than supervised transfer learning and shows consistent performance across demographics (no significant disparities,\n                      <jats:italic>p<\/jats:italic>\n                      &amp;gt; 0.05; several subgroup comparisons are statistically underpowered). New analyses in this revision include Monte Carlo Dropout uncertainty quantification, robustness evaluation under acquisition perturbations, and a computational-complexity assessment.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>This study provides preliminary evidence that ethically developed pain assessment AI\u2013trained without exploiting vulnerable populations\u2013can achieve clinically useful performance, and proposes a methodological pathway toward this goal that warrants confirmation through prospective, adequately powered clinical validation.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/frai.2026.1827727","type":"journal-article","created":{"date-parts":[[2026,8,7]],"date-time":"2026-08-07T05:28:48Z","timestamp":1786080528000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Zero-shot multimodal pain estimation via synthetic pain simulation and domain-invariant learning"],"prefix":"10.3389","volume":"9","author":[{"given":"Oussama","family":"El Othmani","sequence":"first","affiliation":[{"name":"Computer Science Department, Military Academy of Fondouk Jedid","place":["Nabeul, Tunisia"]},{"name":"Military Research Center","place":["Aouina, Tunisia"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sami","family":"Naouali","sequence":"additional","affiliation":[{"name":"Information Systems Department, College of Computer Science and Information Technology, King Faisal University","place":["Al Ahsa, Saudi Arabia"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,7]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"585","DOI":"10.1007\/s41060-024-00649-z","article-title":"Decoding pain: prediction under different emotional contexts through physiological signals","volume":"19","author":"Alves","year":"2025","journal-title":"Int. J. Data Sci. Anal"},{"key":"B2","doi-asserted-by":"publisher","first-page":"1731812","DOI":"10.3389\/fncom.2026.1731812","article-title":"Metaheuristic-driven dual-layer model for classifying Alzheimer's disease stages","volume":"20","author":"Anicin","year":"2026","journal-title":"Front. Comput. Neurosci"},{"key":"B3","doi-asserted-by":"publisher","first-page":"484","DOI":"10.1007\/978-3-031-64836-6_47","article-title":"\u201cDeep learning-based pain intensity estimation from facial expressions,\u201d","author":"Ben Aoun","year":"2023","journal-title":"International conference on intelligent systems design and applications"},{"key":"B4","doi-asserted-by":"publisher","first-page":"104688","DOI":"10.1016\/j.imavis.2023.104688","article-title":"Synthetic data for face recognition: current state and future prospects","volume":"135","author":"Boutros","year":"2023","journal-title":"Image Vis. Comput"},{"key":"B5","doi-asserted-by":"publisher","first-page":"211","DOI":"10.1016\/j.artmed.2004.12.003","article-title":"Machine recognition and representation of neonatal facial displays of acute pain","volume":"36","author":"Brahnam","year":"2006","journal-title":"Artif. Intell. Med"},{"key":"B6","first-page":"1597","article-title":"\u201cA simple framework for contrastive learning of visual representations,\u201d","volume-title":"International conference on machine learning","author":"Chen","year":"2020"},{"key":"B7","article-title":"Ilvr: conditioning method for denoising diffusion probabilistic models","author":"Choi","year":"2021","journal-title":"arXiv preprint arXiv:2108.02938"},{"key":"B8","doi-asserted-by":"publisher","DOI":"10.1109\/ICRA46639.2022.9811954","article-title":"\u201cSelf-supervised representation learning for reliable robotic monitoring of fruit anomalies,\u201d","author":"Choi","year":"2022","journal-title":"2022 international conference on robotics and automation (ICRA)"},{"key":"B9","doi-asserted-by":"publisher","first-page":"232","DOI":"10.1097\/00002508-199609000-00011","article-title":"Barriers to optimal pain management in infants, children, and adolescents social barriers to optimal pain management in infants and children","volume":"12","author":"Craig","year":"1996","journal-title":"Clin. J. Pain"},{"key":"B10","first-page":"8780","article-title":"\u201cDiffusion models beat gans on image synthesis,\u201d","author":"Dhariwal","year":"2021","journal-title":"Advances in neural information processing systems"},{"key":"B11","article-title":"An image is worth 16x16 words: transformers for image recognition at scale","author":"Dosovitskiy","year":"2020","journal-title":"arXiv preprint arXiv:2010.11929"},{"key":"B12","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1093\/acprof:oso\/9780195327939.003.0008","article-title":"Lie catching and microexpressions","volume":"1","author":"Ekman","year":"2009","journal-title":"Philos. Decept"},{"key":"B13","doi-asserted-by":"publisher","author":"El Othmani","year":"2025","DOI":"10.21203\/rs.3.rs-7496944\/v1"},{"key":"B14","doi-asserted-by":"publisher","first-page":"1150264","DOI":"10.3389\/fpain.2023.1150264","article-title":"Multimodal physiological sensing for the assessment of acute pain","volume":"4","author":"Fernandez Rojas","year":"2023","journal-title":"Front. Pain Res"},{"key":"B15","first-page":"1","article-title":"Domain-adversarial training of neural networks","volume":"17","author":"Ganin","year":"2016","journal-title":"J. Mach. Learn. Res"},{"key":"B16","doi-asserted-by":"publisher","first-page":"1372814","DOI":"10.3389\/fpain.2024.1372814","article-title":"Multimodal automatic assessment of acute pain through facial videos and heart rate signals utilizing transformer-based architectures","volume":"5","author":"Gkikas","year":"2024","journal-title":"Front. Pain Res"},{"key":"B17","doi-asserted-by":"publisher","first-page":"139","DOI":"10.1145\/3422622","article-title":"Generative adversarial networks","volume":"63","author":"Goodfellow","year":"2020","journal-title":"Commun. ACM"},{"key":"B18","doi-asserted-by":"publisher","first-page":"788","DOI":"10.1109\/TAFFC.2021.3055294","article-title":"Acute stress state classification based on electrodermal activity modeling","volume":"14","author":"Greco","year":"2021","journal-title":"IEEE Trans. Affect. Comput"},{"key":"B19","doi-asserted-by":"publisher","first-page":"e0140330","DOI":"10.1371\/journal.pone.0140330","article-title":"Pain intensity recognition rates via biopotential feature patterns with support vector machines","volume":"10","author":"Gruss","year":"2015","journal-title":"PLoS ONE"},{"key":"B20","doi-asserted-by":"publisher","first-page":"1216","DOI":"10.1016\/S1474-4422(14)70103-6","article-title":"Pain assessment in elderly adults with dementia","volume":"13","author":"Hadjistavropoulos","year":"2014","journal-title":"Lancet Neurol"},{"key":"B21","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00975","article-title":"\u201cMomentum contrast for unsupervised visual representation learning,\u201d","author":"He","year":"2020","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"key":"B22","first-page":"6840","article-title":"\u201cDenoising diffusion probabilistic models,\u201d","author":"Ho","year":"2020","journal-title":"Advances in neural information processing systems"},{"key":"B23","article-title":"Classifier-free diffusion guidance","author":"Ho","year":"2022","journal-title":"arXiv preprint arXiv:2207.12598"},{"key":"B24","doi-asserted-by":"publisher","first-page":"305","DOI":"10.1109\/RBME.2025.3632213","article-title":"FHIR in focus: enabling biomedical data harmonization for intelligent healthcare systems","volume":"19","author":"Hornback","year":"2025","journal-title":"IEEE Rev. Biomed. Eng"},{"key":"B25","doi-asserted-by":"publisher","first-page":"854","DOI":"10.1109\/JSTSP.2016.2535962","article-title":"Methods for person-centered continuous pain intensity assessment from bio-physiological channels","volume":"10","author":"K\u00e4chele","year":"2016","journal-title":"IEEE J. Sel. Top. Signal Process"},{"key":"B26","doi-asserted-by":"publisher","first-page":"3122","DOI":"10.1109\/TBME.2021.3065218","article-title":"Sensitive physiological indices of pain based on differential characteristics of electrodermal activity","volume":"68","author":"Kong","year":"2021","journal-title":"IEEE Trans. Biomed. Eng"},{"key":"B27","doi-asserted-by":"publisher","first-page":"155","DOI":"10.1016\/j.jcm.2016.02.012","article-title":"A guideline of selecting and reporting intraclass correlation coefficients for reliability research","volume":"15","author":"Koo","year":"2016","journal-title":"J. Chiropr. Med"},{"key":"B28","article-title":"Pain in 3D: generating controllable synthetic faces for automated pain assessment","author":"Lin","year":"2026","journal-title":"arXiv preprint arXiv:2509.16727"},{"key":"B29","doi-asserted-by":"publisher","first-page":"334","DOI":"10.3389\/fnhum.2017.00334","article-title":"Improving eeg-based emotion classification using conditional transfer learning","volume":"11","author":"Lin","year":"2017","journal-title":"Front. Hum. Neurosci"},{"key":"B30","doi-asserted-by":"publisher","DOI":"10.1109\/ACIIW.2019.8925076","article-title":"\u201cPain detection with fnirs-measured brain signals: a personalized machine learning approach using the wavelet transform and bayesian hierarchical modeling with dirichlet process priors,\u201d","author":"Lopez-Martinez","year":"2019","journal-title":"2019 8th international conference on affective computing and intelligent interaction workshops and demos (ACIIW)"},{"key":"B31","doi-asserted-by":"publisher","DOI":"10.1109\/FG.2011.5771462","article-title":"\u201cPainful data: the UNBC-McMaster shoulder pain expression archive database,\u201d","author":"Lucey","year":"2011","journal-title":"2011 IEEE international conference on automatic face"},{"key":"B32","doi-asserted-by":"publisher","first-page":"57","DOI":"10.1007\/s11063-025-11735-z","article-title":"Parkinson's detection from gait time series classification using modified metaheuristic optimized long short term memory","volume":"14","author":"Markovic","year":"2025","journal-title":"Neural Proc. Lett"},{"key":"B33","doi-asserted-by":"publisher","first-page":"106969","DOI":"10.1016\/j.applanim.2026.106969","article-title":"Time budget analysis with large vision language models for automated pain assessment in horses","volume":"299","author":"Martvel","year":"2026","journal-title":"Appl. Animal Behav. Sci"},{"key":"B34","doi-asserted-by":"publisher","first-page":"e0001301","DOI":"10.1371\/journal.pdig.0001301","article-title":"Avpenet: pain estimation from audio-visual fusion of non-speech sounds","volume":"5","author":"Naouali","year":"2026","journal-title":"PLoS Dig. Health"},{"key":"B35","doi-asserted-by":"publisher","DOI":"10.1109\/ACII.2015.7344578","article-title":"\u201cPain level recognition using kinematics and muscle activity for physical rehabilitation in chronic pain,\u201d","author":"Olugbade","year":"2015","journal-title":"2015 international conference on affective computing and intelligent interaction (ACII)"},{"key":"B36","doi-asserted-by":"publisher","first-page":"4992","DOI":"10.3390\/s22134992","article-title":"An automatic system for continuous pain intensity monitoring based on analyzing data from uni-, bi-, and multi-modality","volume":"22","author":"Othman","year":"2022","journal-title":"Sensors"},{"key":"B37","doi-asserted-by":"publisher","first-page":"479","DOI":"10.3390\/s20020479","article-title":"Innovations in electrodermal activity data collection and signal processing: a systematic review","volume":"20","author":"Posada-Quintero","year":"2020","journal-title":"Sensors"},{"key":"B38","doi-asserted-by":"publisher","first-page":"267","DOI":"10.1016\/j.pain.2008.04.010","article-title":"The structure, reliability and validity of pain expression: evidence from patients with shoulder pain","volume":"139","author":"Prkachin","year":"2008","journal-title":"Pain"},{"key":"B39","first-page":"8748","article-title":"\u201cLearning transferable visual models from natural language supervision,\u201d","volume-title":"International conference on machine learning","author":"Radford","year":"2021"},{"key":"B40","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-87386-7_10","article-title":"\u201cAugmentation and substitution of medical training data with generative adversarial networks for machine learning,\u201d","author":"Radomirovic","year":"2024","journal-title":"International conference on modelling and development of intelligent systems"},{"key":"B41","doi-asserted-by":"publisher","first-page":"1091","DOI":"10.1002\/ajh.25117","article-title":"Multiple myeloma: 2018 update on diagnosis, risk-stratification, and management","volume":"93","author":"Rajkumar","year":"2018","journal-title":"Am. J. Hematol"},{"key":"B42","doi-asserted-by":"publisher","first-page":"3314","DOI":"10.1109\/TCYB.2017.2662199","article-title":"Deep pain: exploiting long short-term memory networks for facial expression classification","volume":"52","author":"Rodriguez","year":"2017","journal-title":"IEEE Trans. Cybern"},{"key":"B43","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01042","article-title":"\u201cHigh-resolution image synthesis with latent diffusion models,\u201d","author":"Rombach","year":"2022","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"key":"B44","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1833","article-title":"\u201cLaion-5b: an open large-scale dataset for training next generation image-text models,\u201d","author":"Schuhmann","year":"2022","journal-title":"Advances in neural information processing systems"},{"key":"B45","first-page":"596","article-title":"\u201cFixmatch: simplifying semi-supervised learning with consistency and confidence,\u201d","author":"Sohn","year":"2020","journal-title":"Advances in neural information processing systems"},{"key":"B46","doi-asserted-by":"publisher","first-page":"CD001069","DOI":"10.1002\/14651858.CD001069.pub5","article-title":"Sucrose for analgesia in newborn infants undergoing painful procedures","volume":"7","author":"Stevens","year":"2016","journal-title":"Cochr. Datab. System. Rev"},{"key":"B47","doi-asserted-by":"publisher","first-page":"839","DOI":"10.3390\/s20030839","article-title":"Two-stream attention network for pain recognition from video sequences","volume":"20","author":"Thiam","year":"2020","journal-title":"Sensors"},{"key":"B48","doi-asserted-by":"publisher","DOI":"10.1109\/CYBConf.2013.6617456","article-title":"\u201cThe biovid heat pain database data for the advancement and systematic validation of an automated pain recognition system,\u201d","author":"Walter","year":"2013","journal-title":"2013 IEEE international conference on cybernetics (CYBCO)"},{"key":"B49","doi-asserted-by":"publisher","first-page":"530","DOI":"10.1109\/TAFFC.2019.2946774","article-title":"Automatic recognition methods supporting pain assessment: a survey","volume":"13","author":"Werner","year":"2019","journal-title":"IEEE Trans. Affect. Comput"},{"key":"B50","doi-asserted-by":"publisher","first-page":"162","DOI":"10.1007\/978-3-030-12738-1_12","article-title":"\u201cAutomated pain detection in facial videos of children using human-assisted transfer learning,\u201d","author":"Xu","year":"2019","journal-title":"International workshop on artificial intelligence in health"},{"key":"B51","doi-asserted-by":"publisher","first-page":"101552","DOI":"10.1016\/j.media.2019.101552","article-title":"Generative adversarial network in medical imaging: a review","volume":"58","author":"Yi","year":"2019","journal-title":"Med. Image Anal"},{"key":"B52","doi-asserted-by":"publisher","first-page":"28","DOI":"10.1109\/TAFFC.2019.2926710","article-title":"A comprehensive and context-sensitive neonatal pain assessment using computer vision","volume":"13","author":"Zamzmi","year":"2019","journal-title":"IEEE Trans. Affect. Comput"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1827727\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,7]],"date-time":"2026-08-07T05:28:49Z","timestamp":1786080529000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1827727\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,7]]},"references-count":52,"alternative-id":["10.3389\/frai.2026.1827727"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1827727","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,7]]},"article-number":"1827727"}}