Self-Supervised Contrastive Learning for Medical Time Series: A Systematic Review
Open accessThe DOI matches its Crossref record; 131 of the 132 references it cites were found in a registry or an index, and 1 are not checked yet.
Abstract
Medical time series are sequential data collected over time that measures health-related signals, such as electroencephalography (EEG), electrocardiography (ECG), and intensive care unit (ICU) readings. Analyzing medical time series and identifying the latent patterns and trends that lead to uncovering highly valuable insights for enhancing diagnosis, treatment, risk assessment, and disease progression. However, data mining in medical time series is heavily limited by the sample annotation which is time-consuming and labor-intensive, and expert-depending. To mitigate this challenge, the emerging self-supervised contrastive learning, which has shown great success since 2020, is a promising solution. Contrastive learning aims to learn representative embeddings by contrasting positive and negative samples without the requirement for explicit labels. Here, we conducted a systematic review of how contrastive learning alleviates the label scarcity in medical time series based on PRISMA standards. We searched the studies in five scientific databases (IEEE, ACM, Scopus, Google Scholar, and PubMed) and retrieved 1908 papers based on the inclusion criteria. After applying excluding criteria, and screening at title, abstract, and full text levels, we carefully reviewed 43 papers in this area. Specifically, this paper outlines the pipeline of contrastive learning, including pre-training, fine-tuning, and testing. We provide a comprehensive summary of the various augmentations applied to medical time series data, the architectures of pre-training encoders, the types of fine-tuning classifiers and clusters, and the popular contrastive loss functions. Moreover, we present an overview of the different data types used in medical time series, highlight the medical applications of interest, and provide a comprehensive table of 51 public datasets that have been utilized in this field. In addition, this paper will provide a discussion on the promising future scopes such as providing guidance for effective augmentation design, developing a unified framework for analyzing hierarchical time series, and investigating methods for processing multimodal data. Despite being in its early stages, self-supervised contrastive learning has shown great potential in overcoming the need for expert-created annotations in the research of medical time series.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures





Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Liu et al. [33] provide a comprehensive summary of the various augmentations applied to medical time series data, the architectures of pre-training encoders, the types of fine-tuning classifiers and clusters, and the popular contrastive loss functions.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [33]
The cited source says what the claim attributes to it.
In the citing paper
“The taxonomies proposed by Eldele et al. [19], Deldari et al. [20] and Liu et al. [33] are somewhat similar to our proposed taxonomy, i.e., three taxonomies involve time series self-supervised contrastive learning methods.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [33]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
The PRISMA 2020 statement: An updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · BMJ · doi:10.1136/bmj.n71 · Cited by 108,283
Goldberger, A. L., Amaral, L. A. N., Glass, L., Hausdorff, J. M., Ivanov, P. C., Mark, R. G., Mietus, J. E., Moody, G. B., Peng, C.-K., & Stanley, H. E. (2000) · Circulation · doi:10.1161/01.cir.101.23.e215 · Cited by 14,809
Momentum Contrast for Unsupervised Visual Representation Learning
He, K., Fan, H., Wu, Y., Xie, S., & Girshick, R. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00975 · Cited by 13,347
A simple framework for contrastive learning of visual representations
Chen, T., Kornblith, S., Norouzi, M., & Hinton, G. (2020) · arXiv · doi:10.48550/arxiv.2002.05709 · Cited by 10,857 · cited 6×
“Self-supervised contrastive learning drawn much attention since the development of SimCLR [16] in 2020.”
BackgroundInfluentialMIMIC-III, a freely accessible critical care database
Johnson, A. E. W., Pollard, T. J., Shen, L., Lehman, L.-W. H., Feng, M., Ghassemi, M., Moody, B., Szolovits, P., Anthony Celi, L., & Mark, R. G. (2016) · Scientific Data · doi:10.1038/sdata.2016.35 · Cited by 8,141 · cited 1×
“) [78] for surgical adverse events forecasting.”
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J., Bossuyt, P., Boutron, I., Hoffmann, T., mulrow, C., Shamseer, L., Tetzlaff, J., Akl, E., Brennan, S. E., et al. (2020) · doi:10.31222/osf.io/v7gm2_v1 · Cited by 8,090
“Librispeech: an ASR corpus based on public domain audio books,”
Panayotov, V., Chen, G., Povey, D., & Khudanpur, S. (2015) · 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2015.7178964 · Cited by 6,072
The PRISMA 2020 statement: An updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · PLoS Medicine · doi:10.1371/journal.pmed.1003583 · Cited by 5,148
The impact of the MIT-BIH Arrhythmia Database
Moody, G. B., & Mark, R. G. (2001) · IEEE Engineering in Medicine and Biology Magazine · doi:10.1109/51.932724 · Cited by 4,812
Hannun, A. Y., Rajpurkar, P., Haghpanahi, M., Tison, G. H., Bourn, C., Turakhia, M. P., & Ng, A. Y. (2019) · Nature Medicine · doi:10.1038/s41591-018-0268-3 · Cited by 3,148
Research works printed but not linked: 13
- Spathis D. Perez-Pozuelo I. Brage S. Wareham N.J. Mascolo C. Learning generalizable physiological representations from large-scale wearable data arXiv 2020 2011.04601
- Pan L. Feng Z. Peng S. A review of machine learning approaches, challenges and prospects for computational tumor pathology arXiv 2022 2206.01728
- Kingma D.P. Welling M. Auto-encoding variational bayes arXiv 2013 1312.6114
- Wagh N. Wei J. Rawal S. Berry B. Barnard L. Brinkmann B. Worrell G. Jones D. Varatharajah Y. Domain-guided Self-supervision of EEG Data Improves Downstream Classification Performance and Generalizability Proceedings of the Machine Learning for Health, PMLR Virtual 6–7 August 2021 130 142
- Ho T.K.K. Armanfard N. Self-Supervised Learning for Anomalous Channel Detection in EEG Graphs: Application to Seizure Analysis arXiv 2022 2208.07448
- Cheng J.Y. Goh H. Dogrusoz K. Tuzel O. Azemi E. Subject-aware contrastive learning for biosignals arXiv 2020 2007.04871
- Huijben I.A. Nijdam A.A. Overeem S. van Gilst M.M. van Sloun R.J. SOM-CPC: Unsupervised Contrastive Learning with Self-Organizing Maps for Structured Representations of High-Rate Time Series arXiv 2022 2205.15875
- Zhang X. Zhao Z. Tsiligkaridis T. Zitnik M. Self-supervised contrastive pre-training for time series via time-frequency consistency arXiv 2022 2206.08496
- Weatherhead A. Greer R. Moga M.A. Mazwi M. Eytan D. Goldenberg A. Tonekaboni S. Learning Unsupervised Representations for ICU Timeseries Proceedings of the Conference on Health, Inference, and Learning, PMLR Virtual 7–8 April 2022 152 168
- Wever F. Keller T.A. Symul L. Garcia V. As easy as APC: Overcoming missing data and class imbalance in time series with self-supervised learning arXiv 2021 2106.15577
- Faltys M. Zimmermann M. Lyu X. Hüser M. Hyland S. Rätsch G. Merz T. HiRID, a high time-resolution ICU dataset (version 1.1. 1) PhysioNet 2021 Available online: https://physionet.org/content/hirid/1.1.1/ (accessed on 18 February 2021) 10.13026/nkwc-js72 · DOI
- van den Oord A. Li Y. Vinyals O. Representation learning with contrastive predictive coding arXiv 2018 1807.03748
- Rocha B. Filos D. Mendes L. Vogiatzis I. Perantoni E. Kaimakamis E. Natsiavas P. Oliveira A. Jácome C. Marques A. A respiratory sound database for the development of automated classification Proceedings of the International Conference on Biomedical and Health Informatics Springer Berlin/Heidelberg, Germany 2018 33 37
Cited by
Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects
Zhang, K., Wen, Q., Zhang, C., Cai, R., Jin, M., Liu, Y., Zhang, J. Y., Liang, Y., Pang, G., Song, D., et al. (2024) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2024.3387317 · Cited by 295
“The second category focuses on the key components of time series modeling based on deep neural networks, such as data augmentation [33], [45], [46], [47], model structure [33], [48], [49].”
BackgroundMethodInfluentialTahir, M. A. A., Norouzi, M., Khan, S. S., Davie, J., Yamanaka, S., & Ashraf, A. B. (2024) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2024.109302 · Cited by 56
Applications of Self-Supervised Learning to Biomedical Signals: A Survey
Pup, F. D., & Atzori, M. (2023) · IEEE Access · doi:10.1109/access.2023.3344531 · Cited by 43
“In fact, previously cited works focus on different types of data (medical imaging) [19], [20], specific biomedical signals (EEG) [21], or specific self-supervised learning paradigms (contrastive learning) [15].”
BackgroundLarge language models for bioinformatics
Ruan, W., Lyu, Y., Zhang, J., Cai, J., Shu, P., Yang, G., Lu, Y., Gao, S., Wang, Y., Wang, P., et al. (2025) · Quantitative Biology · doi:10.1002/qub2.70014 · Cited by 48
“Language models are trained using self ‐ supervised learning methods [85].”
Self-citationTS-MAE: A masked autoencoder for time series representation learning
Liu, Q., Ye, J., Liang, H., Sun, L., & Du, B. (2024) · Information Sciences · doi:10.1016/j.ins.2024.121576 · Cited by 20
Self-Supervised Gait Event Detection from Smartphone IMUs for Human Performance and Sports Medicine
TUDOR, A. C., Mănescu, D. C., & Mănescu, D. C. (2025) · Applied Sciences · doi:10.3390/app152211974 · Cited by 18
“Prior work has highlighted the importance of augmentation choice but often without systematic ablation, leaving uncertainty about which transformations most directly enhance cross-dataset generalization [37–40].”
Universal representations in cardiovascular ECG assessment: A self-supervised learning approach
Liu, Z., Lin, C., Hsu, Y.-C., Chen, J., Chang, P., Wen, M., & Kuo, C. (2024) · International Journal of Medical Informatics · doi:10.1016/j.ijmedinf.2024.105742 · Cited by 14
Self-citationFADE: Forecasting for anomaly detection on ECG
Ruiz-Barroso, P., Castro, F. M., Miranda, J., Constantinescu, D.-A., Atienza, D., & Guil, N. (2025) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2025.108780 · Cited by 13
Deep Conv-LSTM Network for Arrhythmia Detection using ECG Data
Mukhametkaly, A., Momynkulov, Z., Kurmanbekkyzy, N., & Omarov, B. S. (2023) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2023.0140973 · Cited by 12
Data mining and machine learning in HIV infection risk research: An overview and recommendations
Ge, Q., Lu, X., Jiang, R., Zhang, Y., & Zhuang, X. (2024) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2024.102887 · Cited by 13
SleepFC: Feature Pyramid and Cross-Scale Context Learning for Sleep Staging
Li, W., Liu, T., Xu, B., & Song, A. (2024) · IEEE Transactions on Neural Systems and Rehabilitation Engineering · doi:10.1109/tnsre.2024.3406383 · Cited by 16
“The laboring process is empirical and time-consuming [4], [5].”
BackgroundA Review on the Applications of GANs for 3D Medical Image Analysis
Usama, Z., Alavi, A., & Chan, J. (2025) · Applied Sciences · doi:10.3390/app152011219 · Cited by 15
Self-citationMulti-scale contrast approach for stock index prediction with adaptive stock fusion
Gao, J., Wang, S., He, C., & Qin, C. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.125590 · Cited by 10
Intelligent Clinical Decision Support System for Managing COPD Patients
Pereira, J., Antunes, N., Rosa, J., Ferreira, J. C., Mogo, S., & Pereira, M. (2023) · Journal of Personalized Medicine · doi:10.3390/jpm13091359 · Cited by 12
“These models are able to identify patterns and trends from sequential data collected over time, such as health-related signals [25,26].”
Early stage lung cancer detection from speech sounds in natural environments
Ankışhan, H., Ulucanlar, H., Aktürk, İ., Kavak, K. A., Bağcı, U., & Yenigün, B. M. (2024) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2024.106628 · Cited by 9
Evaluating AI-powered predictive solutions for MRI in lumbar spinal stenosis: a systematic review
Al–antari, M. A., Salem, S., Raza, M., Elbadawy, A. S., Bütün, E., Aydın, A. A., Aydoğan, M., Ertuğrul, B., Talo, M., & Gu, Y. H. (2025) · Artificial Intelligence Review · doi:10.1007/s10462-025-11185-y · Cited by 11
A self-supervised multimodal framework for 1D physiological data fusion in remote health monitoring
Cañellas, M. L., Casado, C. Á., Nguyen, L. N., & López, M. B. (2025) · Information Fusion · doi:10.1016/j.inffus.2025.103397 · Cited by 13
Contrastive Learning for Network Intrusion Detection: A Comprehensive Survey
Tan, X., Cheng, J., Li, H., & Yang, Y. (2024) · International Conference on Computer, Internet of Things and Smart City (CIoTSC) · doi:10.1145/3731867.3731894 · Cited by 10
“[31]provideareviewontheapplicationofself-supervised contrastive learning to medical time series.”
Semi-Supervised End-To-End Contrastive Learning For Time Series Classification
Cai, H., & Zhang, X. (2026) · Neural Processing Letters · doi:10.1007/s11063-026-11869-8 · Cited by 9
“We measure the supervised contrastive loss based on representation z l : The z l + is a representation derived from a sample that belongs to the same class as z l , i.e., y = y + .”
Self-citationAutomated Contrastive Learning Strategy Search for Time Series
Jing, B., Wang, Y., Sui, G., Hong, J., He, J., Yang, Y., Li, D., & Ren, K. (2024) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3627673.3680086 · Cited by 14
Self-Supervised Representation Learning for EEG-Based Detection of Neurodegenerative Diseases
Pup, F. D., Tshimanga, L. F., Zanola, A., Taffarello, L., Tentori, E., & Atzori, M. (2025) · Applied Sciences · doi:10.3390/app152413275 · Cited by 8
Effective Anomaly Detection in 5G Networks via Transformer-Based Models and Contrastive Learning
Sheikhi, S., Kostakos, P., & Pirttikangas, S. (2024) · Cyber Security in Networking Conference (CSNet) · doi:10.1109/csnet64211.2024.10851755 · Cited by 7
“Contrastive learning, a type of self-supervised learning, has gained attention for its ability to learn representations by distinguishing between similar and dissimilar pairs of data [5], [6].”
Liu, W., Pan, S., Chang, S., Huang, Q., & Jiang, N. (2025) · Applied Soft Computing · doi:10.1016/j.asoc.2025.112871 · Cited by 8
Cui, K., Li, J., Liu, Y., Zhang, X., Hu, Z., & Wang, M. (2025) · IEEE Transactions on Computational Social Systems · doi:10.1109/tcss.2025.3602913 · Cited by 12
“Typically, contrastive learning methods involve pretraining representations on relatively large datasets, which are then fine-tuned for downstream tasks [34].”
Self-citationSoumma, S. B., Peterson, D., Mehta, S. H., & Ghasemzadeh, H. G. (2026) · ACM Transactions on Computing for Healthcare · doi:10.1145/3802589 · Cited by 9
“Additionally, the models struggle with high false-positive rates, which can reduce their effectiveness in real-world settings where minimizing such errors is crucial for patient safety and usability.”
Lee, J., Ham, I., Kim, Y., & Ko, H. (2024) · Sensors · doi:10.3390/s24247932 · Cited by 6
“This approach effectively captures nuanced patterns and relationships within the data, enhancing the model’s capacity to generalize in downstream tasks [20].”
Multimodal cardiovascular risk profiling using self-supervised learning of polysomnography
He, Z., Li, H., Yuan, G., Killgore, W. D. S., Quan, S. F., Chen, C. X., & Li, A. (2025) · SLEEP · doi:10.1093/sleep/zsaf371 · Cited by 6
HeartBERT: A self-supervised ECG embedding model for efficient and effective medical signal analysis
Tahery, S., Akhlaghi, F. H., Amirsoleimani, T., & Farzi, S. (2026) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2026.111712 · Cited by 5
Liu, W., Pan, S., Li, Z., Chang, S., Huang, Q., & Jiang, N. (2024) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2024.108452 · Cited by 5
Unsupervised Contrastive Learning for Time Series Data Clustering
Bo, C., Xing, Q., Yang, K., Wu, X., & Li, L. (2025) · Electronics · doi:10.3390/electronics14081660 · Cited by 4
Arabi, H., & Zaidi, H. (2024) · Journal of Imaging Informatics in Medicine · doi:10.1007/s10278-024-01159-x · Cited by 5
Multi-Modal Contrastive Learning for Online Clinical Time-Series ApplicationsPreprint
Baldenweg, F., Burger, M., Rätsch, G., & Kuznetsova, R. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2403.18316 · Cited by 7
“They often rely on InfoNCE (van den Oord et al., 2019) or a variant thereof to pull augmented views of the same data closer together and push different samples apart in the representation space Liu et al. (2023b).”
BackgroundLian, L. (2024) · International Conference on Software, Systems and Information Technology (SSITCON) · doi:10.1109/ssitcon62437.2024.10795934 · Cited by 3
“This research provides the data is directly extracted from Kaggle depository Philosophy Data Project (PDP) collected data affording [16].”
Personalization of AI Using Personal Foundation Models Can Lead to More Precise Digital Therapeutics
Washington, P. Y. (2025) · JMIR AI · doi:10.2196/55530 · Cited by 6
“Another approach is to perform contrastive learning algorithms such as SimCLR [22] on the signals to maximize representational similarity between augmented versions of the same time period while minimizing similarity between 2 distinct time windows (Figure 2B) [23,24].”
Hassan, S. U., Adil, M., Khan, N. A., Khan, M. K., & Jin, Z. (2026) · IEEE Internet of Things Journal · doi:10.1109/jiot.2026.3695869 · Cited by 3
“…and large unlabeled corpora analogous to ImageNet or PhysioNet will be required to establish stable encoder families for vital-sign sensing [82]. a) Physics-guided and hybrid architectures: Physics-guided learning is one of the most radar-specific directions in mmWave health monitoring…”
A survey on medical multimodal retrieval-augmented generation (RAG)
Wang, R., Wang, J., Huang, Y., Guo, H., & Pang, J. (2026) · High-Confidence Computing · doi:10.1016/j.hcc.2026.100415 · Cited by 3
Emerging trends and clinical challenges in AI-enhanced emotion diagnosis using physiological data
Tsai, Y.-Y., Wu, G., Chen, Y.-J., Lin, K., Wu, J., Hsu, C., & Liao, L. (2025) · Medical & Biological Engineering & Computing · doi:10.1007/s11517-025-03435-6 · Cited by 3
Lou, B., Kan, K., Liu, H., Feng, R., Zhang, X., Yuan, Z., Zhang, L., & She, J. (2024) · Reviews in Cardiovascular Medicine · doi:10.31083/j.rcm2505151 · Cited by 3
“Supervised learning has emerged as a particularly apt approach for tasks characterized by well-defined objectives, given its amenability to facile quantification through diverse metrics, thereby facilitating the straightforward evaluation of accuracy and efficacy [33,34].”
BackgroundSelf-citationMugambi, L., Maina, C. W., & Zuhlke, L. J. (2025) · Journal of Imaging · doi:10.3390/jimaging11040097 · Cited by 3
“However, contrastive learning offers a solution by learning from positive and negative samples without explicit labels, alleviating label scarcity challenges [24].”
Data-Efficient Sleep Staging with Synthetic Time Series Pretraining
Grieger, N., Mehrkanoon, S., & Bialonski, S. (2025) · Algorithms · doi:10.3390/a18090580 · Cited by 2
“A variant of this idea is self-supervised learning [32, 2], which allows neural networks to be pretrained on large and heterogeneous datasets without explicitly labeled examples.”
Özel, E. (2026) · IEEE Access · doi:10.1109/access.2026.3697731 · Cited by 4
“Deep learning consistently outperforms traditional machine learning in modern gait analysis [6].”
Smart Attendance Management using a Self-Supervised Learning Approach
Vikram, D., Ajina, H., Gokilapriya, S., & Malini, M. (2024) · International Conference on Image Processing and Capsule Networks (ICIPCN) · doi:10.1109/icipcn63822.2024.00032 · Cited by 2
“More accurate attendance tracking and anomaly identification are made possible by the system's ability to extract high-dimensional features from raw attendance data through SSL-driven feature extraction [11] [12].”
Ren, H., Luo, Z., Jing, F., Zhang, X., He, H., Yu, Y., & Zhao, D. (2024) · IEEE International Conference on Data Science in Cyberspace · doi:10.1109/dsc63484.2024.00106 · Cited by 2
“Moreover, this approach facilitates in-depth analysis of multimodal medical data through feature space alignment techniques [35] and demonstrates potential in addressing the scarcity of labels in medical time series analysis [36].”
Self-citationOptimized Contrastive Transformer Models for Self-Supervised 5G Network Intrusion Detection
Sheikhi, S., Kostakos, P., Pirttikangas, S., & Lovén, L. (2025) · International Generative AI and Computational Language Modelling Conference (GACLM) · doi:10.1109/gaclm67198.2025.11231874 · Cited by 2
“Asaself-supervised paradigm, contrastive learning has gained significant interest due to its ability to derive robust representations by contrasting positive (similar) and negative (dissimilar) data pairs [5], [6].”
Instruction-based Time Series Editing
Qiu, J., Guo, D., Sullivan, B., Henry, T. R., & Hartvigsen, T. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770854.3780299 · Cited by 3
“Pioneered by CLIP [42] for image–language alignment, contrastive methods have since been extended to time series for tasks such as retrieval, classification, and forecasting.”
A Novel Data Mining Framework for Discovering Association Rules in Multivariate Time Series
Wang, S., Lyu, J., & Zhang, B. (2024) · International Conference on Advanced Algorithms and Control Engineering (ICAACE) · doi:10.1109/icaace61206.2024.10548882 · Cited by 1
“Time series data consists of observations of specific indicators recorded at regular intervals over a defined period, encompassing various domains such as meteorological data[1], electrocardiogram (ECG) readings[2], and stock market indices.”
BackgroundChen, X., & Hauskrecht, M. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-66538-7_5 · Cited by 1
Wang, S., Zhang, L., Pan, Y., Chen, H., & Pang, Z. (2024) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2024.3510692 · Cited by 1
“Incontrastive learning, we use data augmentation methods to generate different temporal and spatial sequence views of the same trajectory, then obtain their respective representation [20], [21].”
Semantic-Aware Representation of Multi-Modal Data for Data Ingress: A Literature Review
Lamart, P., Yu, Y., & Berger, C. (2024) · Euromicro Conference on Software Engineering and Advanced Applications (SEAA) · doi:10.1109/seaa64295.2024.00026 · Cited by 1
“[10] or Xu et al. [22] also highlight the lack of scalability to multi-modal data and the need to push research in this direction.”
Enhanced Log Anomaly Detection with Contrastive Learning and BERT Embeddings
Cui, M., Gao, T., Xiang, H., Xu, X., Lu, C., Xiong, J., & Xue, S. (2024) · IEEE International Symposium on Parallel and Distributed Processing with Applications (ISPA) · doi:10.1109/ispa63168.2024.00024 · Cited by 1
“…large language models for deep semantic analysis of logs to enrich the contextual information of log events; then, through contrastive learning [13] [31], it further enhances the quality of these semantic representations, optimizing the model’s ability to recognize and differentiate anomaly…”
Time Series Forecasting Based on Temporal Networks Evolution and Dynamic Constraints
Peng, Y., Li, H., & Han, X. (2025) · IEEE Access · doi:10.1109/access.2025.3577499 · Cited by 1
“Furthermore, scientific researchers have proposed new methods to achieve better predictions [27].”
Chui, K. T., Arya, V., Gupta, B. B., Torres-Ruiz, M. J., & Attar, R. W. (2025) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2025.068842 · Cited by 1
“Some typical examples are medical time-series data [25] and medical image segmentation [26].”
Wang, X., Ozek, B., Mohan, A., Ravari, A., Zilbershot, O., & Afghah, F. (2026) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp55912.2026.11461030 · Cited by 1
Zhu, X., Cao, Y., Wang, F., Zhang, J., An, Y., Sun, Y., & Su, S. (2026) · Mechanical Systems and Signal Processing · doi:10.1016/j.ymssp.2026.114574 · Cited by 1
Yang, X., Lv, Z., Lv, H., Zhou, Q., Jiang, W., Ziqun, M. E. N. G., & Yang, W. (2026) · JMIR Medical Informatics · doi:10.2196/82762 · Cited by 1
“Similar supervised contrastive approaches have shown promise in related health care domains, such as mortality prediction [24,25] and patient state representation [26,27], and our work further demonstrates their applicability in the context of real-time warning tasks.”
Ghayvat, H., Tiwari, V., Khare, S., Khan, M. A., Awais, M., Xin, N., Tiwari, P., EI-Sayed, H., & Geddam, R. (2026) · Computer Science Review · doi:10.1016/j.cosrev.2026.101028 · Cited by 1
Contrastive Self-Supervised Learning in Medical Signal Processing: Literature Review and Case Study
Alouani, Z., Daanouni, O., Saleh, S., Moujahid, H., Cherradi, B., & Bouattane, O. (2024) · International Conference on Circuit, Systems and Communication (ICCSC) · doi:10.1109/iccsc62074.2024.10616756 · Cited by 1
“Contrastive SSL is composed of three stages [32], (1 The stages of contrastive SSL are illustrated in figure 1, and explained in the next of this section.”
BackgroundDepression Detection via Contextual Correlation Extractor and Individual-Level Clustering
Zhao, L., Xu, C., Li, W., Hou, H.-R., & Meng, Q. (2025) · Chinese Control Conference (CCC) · doi:10.23919/ccc64809.2025.11179266 · Cited by 1
“Since human physiological signals have typical characteristics of time series signals, researchers combined the study of human physiological signals with contrastive learning methods [10] and achieved good results.”
Data analytics framework for sparse longitudinal structured biomedical data
Sixtová, I., Skopal, T., & Uher, T. (2023) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm58861.2023.10385940
“Another opportunity is to address the sparsity of the data with self-supervised learning representations [33].”
BackgroundHejč, J., Ředina, R., Pospíšil, D., Rakova, I., Kolářová, J., & Stárek, Z. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.04.28.591427
“A comprehensive review by Liu et al. (2023) delves deeper into self-supervised contrastive techniques for medical time series.”
BackgroundLiu, Y., Han, J., Ramnarayan, P., & Faisal, A. A. (2025) · Studies in computational intelligence · doi:10.1007/978-3-032-11999-5_16 · Cited by 1
Revisiting self-supervised contrastive learning for imbalanced classification
Gao, X., Ramli, M. I., Rosli, M. M., Jamil, N., & Ariffin, S. M. Z. S. Z. (2025) · International Journal of Power Electronics and Drive Systems/International Journal of Electrical and Computer Engineering · doi:10.11591/ijece.v15i2.pp1949-1960 · Cited by 3
Madusanka, N., Malekroodi, H. S., Herath, H. M. K. K. M. B., Hewage, C. T. E. R., Yi, M., & Lee, B.-I. (2025) · Journal of Imaging · doi:10.3390/jimaging11070220
“Contrastive learning has shown particular promise in medical audio analysis, with recent studies demonstrating its effectiveness for respiratory sound classification [28], medical time series analysis [29], and general-purpose audio representations [30,31].”
Research on AI Model of Machine Learning in Intelligent Recommendation System
Kong, Y. (2025) · IEEE International Conference on Image Processing and Computer Applications (ICIPCA) · doi:10.1109/icipca65645.2025.11138177
“A self-supervised recommendation model VESPA-CSRM based on Variational Autoencoder (VAE) [1-3] and Contrastive Learning [4-6] is proposed.”
Contrastive Learning Model for Wearable-Based Ataxia Assessment
Lee, J., Oubre, B., Daneault, J., Stephen, C. D., Schmahmann, J. D., Gupta, A. S., & Lee, S. I. (2025) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2025.3608674
“To further reduce this dependency, future work could explore self-supervised contrastive learning approaches, which have shown promise in analyzing medical time-series data [38], [39].”
PHASEY: A Contrastive Learning Approach for Enhanced Human Gait Phases Recognition
Urvashi, U., Kumar, D., Kukreja, V., & Dogra, A. (2025) · Journal of Computer Science · doi:10.3844/jcssp.2025.1795.1810
“The memory bank is usually used to store the features of the training data (Liu et al. , 2023).”
A Survey on Machine Learning Models for Detection of Anomalies in Photovoltaic Systems
Bhari, T., & Zhu, Y. (2025) · Australasian Universities Power Engineering Conference · doi:10.1109/aupec66173.2025.11219517
Arifin, A., Harbi, H., Indriani, A. S., Laila, I., Abdullah, B., Alridho, Idris, I., & Prakosa, J. A. (2025) · Signals · doi:10.3390/signals6040067
“…self-supervised representation learning has proven effective for ECG, EEG, and wearable sensor data, as it reduces the labeling burden while preserving clinically meaningful patterns, whereas unsupervised clustering has been shown to be reliable in profiling large-scale cardiac dynamics [12–14].”
CoGenT: A Unified Contrastive-Generative Framework for Time Series Classification
Liu, Z., Alavi, A., Li, M., & Zhang, X. (2025) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2025.3637168
“Here, we randomly sample 30% of the training set (a label ratio of 0.3, consistent with common practices [32]) for supervised fine-tuning.”
InfluentialSelf-citationKhanh, G. L., Cong, T. D., & Lan, H. H. T. (2025) · The Transport and Communications Science Journal · doi:10.47869/tcsj.77.1.10 · Cited by 2
“These approaches reduce dependency on large labelled datasets and have shown significant promise in time-series domains [10].”
Odeh, V. A., Xiaoling, Z., Idowu, A., Suleman, D., Uko, K., Jackson, T., Yakoi, P. S., & Ssemwogerere, R. (2025) · International Conference on Signal Processing Systems (ICSPS) · doi:10.1109/icsps66615.2025.11347910
Zewail, R. (2026) · IEEE Access · doi:10.1109/access.2026.3658301
“The utilization of these techniques in biosignals is an active area of research [8], [9], [10].”
Exploring Data-Oriented Strategies for Training ECG Self-supervised Models
Silva, R. D. C., Medronha, A., & Silva, D. F. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-15990-8_15
Aortic Valve Disease Screening from PPG via Physiology-Guided Self-Supervised LearningPreprint
Wang, J., Zhao, Q., Chen, Z., Sun, Z., Zhang, D., Zhou, Y., & Hong, S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2602.04266 · Cited by 1
Align and Adapt: Multimodal Multiview Human Activity Recognition Under Arbitrary View Combinations
Nguyen, D., & Le‐Khac, N. (2026) · IEEE Sensors Journal · doi:10.1109/jsen.2026.3718255
“Second, contrastive learning is widely used for representation learning [11], [35].”
Li, Y., Peng, H., Feng, W., Nie, C., Wang, H., Lian, J., & Zhao, Y. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.109774
Wang, M., Shi, Z., Wang, Y., & Lu, J. (2026) · Digital Health · doi:10.1177/20552076261425362
“51,52 Traditional supervised approaches require extensive labeled datasets, which are particularly challenging to obtain for rare lesions like LSTs. 53 Our DINO pretraining effectively leveraged 150,168 unlabeled images to learn robust visual representations, signi fi cantly reducing the annotation…”
A Systematic Literature Review of Semi-Supervised Learning on Time-Series Data
Ahsan, F. M., Ardiyanto, I., & Wibirama, S. (2025) · International Conference on Data Science and Geoinformatics (ICDSG) · doi:10.1109/icdsg67714.2025.11381355
“For example, labeling medical or activity recognition data requires expert and manual effort, limiting the scalability of supervised approaches when unlabeled data are abundant but annotations are scarce [8].”
Contrastive multimodal deep learning for survival prediction in grade 2/3 gliomas
Hua, P., Lin, C.-C., Fenlon, T., Liu, X., & Hassanpour, S. (2026) · JNCI Cancer Spectrum · doi:10.1093/jncics/pkag039
“24,25 It is a representation learning approach that trains models to bring related samples closer together in the embedding space while pushing unrelated samples farther apart, therefore producing more informative, generalizable representations.”
Pang, H., Zhang, H., Zheng, Y., & Fang, H. (2026) · IEEE Sensors Journal · doi:10.1109/jsen.2026.3684713
“…have applied contrastive learning to time-series signals in fields, such as cardiac monitoring [27], bio-signal processing [28], and medical diagnostics [29], these models cannot be directly adapted to the unique dynamics and variability inherent in proprioceptive sensing for robotic systems.”
Heterogeneous multi-objective contrastive learning for stock prediction
Loua, M. A., Ouyang, X., Wang, X., Wang, J., Feng, Y., & Yang, X. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.132492
Self-Supervised Learning for Small Datasets Using Data Augmentation and Contrastive Learning
Deorari, M. C., Rajput, I. S., Kumar, A. S., & Tyagi, S. (2026) · International Conference on Advancing Sustainable Solutions through Technologies (ICASST) · doi:10.1109/icasst68917.2026.11484242
“Back in the day, word2vec and BERT showed how unsupervised learning could work well for text data [7].”
Generalizing Intensive Care AI across Time Scales in Resource-Limited SettingsPreprint
Devadiga, A., Singh, P., Sankar, J., Lodha, R., & Sethi, T. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7224124
Cognitive State and Emotion Recognition Based on Multi-Modal Contrastive Learning for Pre-Training
Li, W., Han, Z., Gao, F., Xing, C., Tang, S., & Guo, Y. (2025) · International Conference on Industrial Automation, Robotics and Control Engineering (IARCE) · doi:10.1109/iarce68366.2025.11485795
Physio-Aware Contrastive Variational Autoencoder for Unsupervised ECG Anomaly Detection
Anem, J., Kalla, K. R., Ankitha, B. S. L., Srinivasarao, P., Lokesh, K., & Ganesh, C. S. (2026) · IEEE International Conference on AI Engineering and Innovations (AIEI) · doi:10.1109/aiei69164.2026.11496894
“In the medical time-series domain, Liu et al. (2023) note a trend where contrastive methods are used to distinguish normal vs abnormal ECGs by embedding beats or segments, learning representations by pulling together augmented views of the same signal and pushing apart others [6].”
A physiology-guided temporal graph learning approach for intensive care mortality prediction
Joseph, J., & Judy, M. V. (2026) · Healthcare Analytics · doi:10.1016/j.health.2026.100467 · Cited by 1
Song, X., Li, Q., Mao, H., & Fan, Y. (2026) · Interdisciplinary Sciences Computational Life Sciences · doi:10.1007/s12539-026-00834-7
EvoGraphConPain: an adaptive multimodal graph intelligence framework for neonatal pain assessment
Othmani, O. E., & Ouersighni, R. (2026) · Machine Vision and Applications · doi:10.1007/s00138-026-01861-6
Rezaee, K., Khosravi, M. B., Rachini, A. S., & Muda, Z. C. (2025) · Asian Conference on Communication and Networks (ASIANComNet) · doi:10.1109/asiancomnet68615.2025.11579627
“Recent reviews covering EEG-specific SSL and also medical time-series more generally confirm broad gains over purely supervised baselines in realistic settings, especially when pretraining is fold-restricted to avoid direct subject-wise information leakage [6–9].”
InfluentialRoy, L., & Sharma, A. (2026) · International Conference on Inventive Computing and Informatics (ICICI) · doi:10.1109/icici68773.2026.11580759
“The result is a significant void in healthcare graph analytics that GNNs have not been able to address, hindering the development of scalable, label-efficient methods [3][4].”
Reducing False Heart Rate Alerts via Uncertainty-Aware Motion-Physiology Modeling
Patil, S., & Deshmukh, P. (2026) · International Conference on Inventive Computing and Informatics (ICICI) · doi:10.1109/icici68773.2026.11581113
“Recent advances in self-supervised and contrastive learning have shown strong transfer to physiological time-series tasks with limited labels [8], [9].”
Time series anomaly detection models with contrastive representation learning
Quijada, F., Drobnyh, K., & Runger, G. C. (2026) · Journal of Intelligent Information Systems · doi:10.1007/s10844-026-01078-z
“Data labeling is often a complex and expensive task, requiring expert-based knowledge to analyze of large bodies of data (Liu et al., 2023; Woodward et al., 2020 ).”
A Review of Contrastive Learning for Stress and Anxiety Detection Using EDA and BVP Signals
Dey, A., & Sethia, D. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-28594-2_27
Karthiga, M., Joseph, E. R., Patil, S., & Saranya, K. (2026) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2026.1854179
MedTVL: Harnessing Vision and Language for Medical Time Series Classification
Ye, J., Li, J., & Tsung, F. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3818883 · Cited by 1
Lee, H., Jeong, S.-M., Seok, C. L., Park, Y., Kim, S., Kim, J. H., Lee, U. C., Jeong, B. H., & Lee, E. C. (2026) · Electronics · doi:10.3390/electronics15163657
Theranostic Smart Bionanosensors for Solid Tumors: Advances, Challenges, and Future Directions
El-Morsy, M. T., Basem, Y., Ali, A. E., Sherif, A., Ata, A., Sabry, F., ElMahsoub, N. M., Doghish, A. S., El-Sayyad, G. S., & Ali, G. A. M. (2026) · Applied Biochemistry and Biotechnology · doi:10.1007/s12010-026-05874-9
Gangemi, S., Manti, S., Virchow, J. C., Jr., & Canonica, G. W. (2026) · Allergy · doi:10.1111/all.70468
Abdelhady, G., Hussein, K. W., & Gad, I. A. A. (2026) · Scientific Reports · doi:10.1038/s41598-026-67450-9 · Cited by 1
Metadata
- Authors
- Liu, Z., Alavi, A., Li, M., & Zhang, X.
- Affiliations
- RMIT University (AU)
- Coles, Melbourne, VIC 3123, Australia
- University of North Carolina at Charlotte (US)
- Corresponding author
- Alavi, Azadeh
- Zhang, Xiang
- Venue
- Sensors
- Publisher
- MDPI AG
- Published
- 2023-04-23
- Volume
- 23
- Issue
- 9
- Pages
- 4221
- ISSN
- 1424-8220 (ISSN-L)
- PMC
- PMC10181273
- OpenAlex
- W4366829097
- Cited by
- 90
- References
- 128
- Citations by year
- 2026: 35 · 2025: 26 · 2024: 25 · 2023: 4
- FWCI
- 1.88
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Sensors (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Figshare (submittedVersion, cc-by)
- Funders
- National Science Foundation (2245894)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning in Healthcare
- Physical Sciences › Computer Science › Signal Processing › Time Series Analysis and Forecasting
- Health Sciences › Medicine › Cardiology and Cardiovascular Medicine › ECG Monitoring and Analysis
- Keywords
- self-supervised contrastive learning, medical time series, limited labeled data, data augmentation, contrastive loss, EEG, ECG, intensive care monitoring, hierarchical time series, multimodal data, medical datasets
- MeSH terms
- Learning, Time Factors, Databases, Factual, Data Mining
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.