Transforming label-efficient decoding of healthcare wearables with self-supervised learning and “embedded” medical domain expertise
Abstract Healthcare wearables are transforming health monitoring, generating vast and complex data in everyday free-living environments. While supervised deep learning has enabled tremendous advances in interpreting such data, it remains heavily dependent on large labeled datasets, which are often d...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-07-01
|
| Series: | Communications Engineering |
| Online Access: | https://doi.org/10.1038/s44172-025-00467-6 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849333502002069504 |
|---|---|
| author | Xiao Gu Zhangdaihong Liu Jinpei Han Jianing Qiu Wenfei Fang Lei Lu Lei Clifton Yuan-Ting Zhang David A. Clifton |
| author_facet | Xiao Gu Zhangdaihong Liu Jinpei Han Jianing Qiu Wenfei Fang Lei Lu Lei Clifton Yuan-Ting Zhang David A. Clifton |
| author_sort | Xiao Gu |
| collection | DOAJ |
| description | Abstract Healthcare wearables are transforming health monitoring, generating vast and complex data in everyday free-living environments. While supervised deep learning has enabled tremendous advances in interpreting such data, it remains heavily dependent on large labeled datasets, which are often difficult and expensive to obtain in clinical practice. Self-supervised contrastive learning (SSCL) provides a promising alternative by learning from unlabeled data, but conventional SSCL frequently overlooks important physiological similarities by treating all non-identical instances as unrelated, which can result in suboptimal representations. In this study, we revisit the enduring value of domain knowledge “embedded” in traditional domain feature engineering pipelines and demonstrate how it can be used to guide SSCL. We introduce a framework that integrates clinically meaningful features—such as heart rate variability from electrocardiograms (ECGs)—into the contrastive learning process. These features guide the formation of more relevant positive pairs through nearest-neighbor matching and promote global structure through clustering-based prototype representations. Evaluated across diverse wearable technologies, our method achieves comparable performance with only 10% labeled data, compared to conventional SSCL approaches with full annotations for fine-tuning. This work highlights the indispensable and sustainable role of domain expertise in advancing machine learning for real-world healthcare, especially for healthcare wearables. |
| format | Article |
| id | doaj-art-9b7d49a2c7d6491f83c20e5d7b1bd5b5 |
| institution | Kabale University |
| issn | 2731-3395 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Communications Engineering |
| spelling | doaj-art-9b7d49a2c7d6491f83c20e5d7b1bd5b52025-08-20T03:45:49ZengNature PortfolioCommunications Engineering2731-33952025-07-014111410.1038/s44172-025-00467-6Transforming label-efficient decoding of healthcare wearables with self-supervised learning and “embedded” medical domain expertiseXiao Gu0Zhangdaihong Liu1Jinpei Han2Jianing Qiu3Wenfei Fang4Lei Lu5Lei Clifton6Yuan-Ting Zhang7David A. Clifton8Department of Engineering Science, University of OxfordDepartment of Engineering Science, University of OxfordBrain and Behaviour Lab, Imperial College LondonDepartment of Biomedical Engineering, The Chinese University of Hong KongDepartment of Engineering Science, University of OxfordSchool of Life Course and Population Sciences, King’s College LondonDepartment of Engineering Science, University of OxfordDepartment of Electronic Engineering, The Chinese University of Hong KongDepartment of Engineering Science, University of OxfordAbstract Healthcare wearables are transforming health monitoring, generating vast and complex data in everyday free-living environments. While supervised deep learning has enabled tremendous advances in interpreting such data, it remains heavily dependent on large labeled datasets, which are often difficult and expensive to obtain in clinical practice. Self-supervised contrastive learning (SSCL) provides a promising alternative by learning from unlabeled data, but conventional SSCL frequently overlooks important physiological similarities by treating all non-identical instances as unrelated, which can result in suboptimal representations. In this study, we revisit the enduring value of domain knowledge “embedded” in traditional domain feature engineering pipelines and demonstrate how it can be used to guide SSCL. We introduce a framework that integrates clinically meaningful features—such as heart rate variability from electrocardiograms (ECGs)—into the contrastive learning process. These features guide the formation of more relevant positive pairs through nearest-neighbor matching and promote global structure through clustering-based prototype representations. Evaluated across diverse wearable technologies, our method achieves comparable performance with only 10% labeled data, compared to conventional SSCL approaches with full annotations for fine-tuning. This work highlights the indispensable and sustainable role of domain expertise in advancing machine learning for real-world healthcare, especially for healthcare wearables.https://doi.org/10.1038/s44172-025-00467-6 |
| spellingShingle | Xiao Gu Zhangdaihong Liu Jinpei Han Jianing Qiu Wenfei Fang Lei Lu Lei Clifton Yuan-Ting Zhang David A. Clifton Transforming label-efficient decoding of healthcare wearables with self-supervised learning and “embedded” medical domain expertise Communications Engineering |
| title | Transforming label-efficient decoding of healthcare wearables with self-supervised learning and “embedded” medical domain expertise |
| title_full | Transforming label-efficient decoding of healthcare wearables with self-supervised learning and “embedded” medical domain expertise |
| title_fullStr | Transforming label-efficient decoding of healthcare wearables with self-supervised learning and “embedded” medical domain expertise |
| title_full_unstemmed | Transforming label-efficient decoding of healthcare wearables with self-supervised learning and “embedded” medical domain expertise |
| title_short | Transforming label-efficient decoding of healthcare wearables with self-supervised learning and “embedded” medical domain expertise |
| title_sort | transforming label efficient decoding of healthcare wearables with self supervised learning and embedded medical domain expertise |
| url | https://doi.org/10.1038/s44172-025-00467-6 |
| work_keys_str_mv | AT xiaogu transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise AT zhangdaihongliu transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise AT jinpeihan transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise AT jianingqiu transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise AT wenfeifang transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise AT leilu transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise AT leiclifton transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise AT yuantingzhang transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise AT davidaclifton transforminglabelefficientdecodingofhealthcarewearableswithselfsupervisedlearningandembeddedmedicaldomainexpertise |