AI and IoT tools for COPD detection and monitoring reviewed across 32 studies

Chronic obstructive pulmonary disease remains a leading cause of illness and death worldwide, and a new systematic literature review argues that artificial intelligence and Internet of Things technologies could reshape how the condition is detected and managed. The review, published in BMC Medical Informatics and Decision Making, synthesizes current evidence on AI- and IoT-enabled approaches for COPD detection and management and identifies the research gaps and implementation challenges that still stand between promising prototypes and routine clinical use.

The authors, Dinuka Bandara, N. M. Hakmanage, M. Chandrasena and R. Tennekoon, conducted the review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. They searched PubMed, IEEE Xplore and Google Scholar for studies published between January 2015 and May 2026. After duplicate removal and a two-stage screening process using predefined eligibility criteria, 32 peer-reviewed studies made it into the final synthesis. The review was registered prospectively, with registration number CRD420261448597 dated 14 July 2026.

Rather than treating COPD technology as a single field, the review organized the selected studies into four domains: early detection, exacerbation prediction, remote monitoring and self-monitoring.

Across the included studies, a consistent picture emerged. AI and IoT technologies can improve COPD detection and management through intelligent analysis of physiological and environmental data collected from wearable devices, smart sensors and digital health platforms. In practice, this means physiological and environmental data collected from wearable devices, smart sensors and digital health platforms can be analyzed by AI and IoT systems rather than assessed only during periodic clinical visits, which the review identifies as a key limitation of conventional approaches to COPD diagnosis and management.

On the algorithmic side, the review reports that traditional machine learning algorithms and deep learning models showed promising performance for COPD detection and exacerbation prediction. Both families of approaches, according to the synthesized studies, performed well enough to be considered promising, although the review stops short of claiming they are ready for unsupervised clinical deployment.

The IoT side of the equation complements the algorithms. IoT-enabled systems supported continuous patient monitoring and self-monitoring in the reviewed literature, extending observation beyond the clinic and giving patients a role in tracking their own condition. The combination is what the authors describe as AI-IoT integration: sensors generate continuous streams of physiological and environmental data, and AI methods convert those streams into detection and prediction outputs that could, in principle, enable continuous, personalized and data-driven healthcare.

Yet the review is equally clear about the weaknesses of the existing evidence base. Most of the 32 studies focused on individual COPD management domains rather than integrated solutions, and the field was limited by small, single-center datasets and inadequate external validation. The review reports that most studies were limited by small, single-center datasets and inadequate external validation, meaning reported performance has not been tested on independent populations.

Those limitations shape the review’s conclusions. The authors conclude that AI-IoT integration has considerable potential to transform COPD management, but they call for future research to prioritize integrated, cost-effective systems that combine early detection, exacerbation prediction, remote monitoring and self-monitoring within a unified framework. They also emphasize that such systems must be supported by large-scale multicenter validation to facilitate real-world clinical adoption, directly addressing the dataset and validation gaps found across the reviewed literature.

For clinicians and researchers, the review functions as a map of a fragmented landscape, in which most studies have focused on individual COPD management domains rather than integrated solutions. The evidence suggests that intelligent analysis of data from wearables, smart sensors and digital health platforms can support COPD detection and exacerbation prediction, and that connected devices can sustain monitoring outside clinical settings. What remains unproven, on the current evidence, is whether these components work reliably together, across diverse populations and health systems, at a cost that makes widespread deployment feasible. The review, published open access on 26 October 2026, is positioned as a synthesis of where the field stands and a statement of what must happen next before AI- and IoT-based COPD tools move from promising studies into everyday care.

Subject of Research: A systematic literature review of artificial intelligence and Internet of Things approaches for COPD detection and management

Article Title: Artificial intelligence and IoT-enabled approaches for COPD detection and monitoring: a systematic literature review

Article References: Bandara, D., Hakmanage, N. M., Chandrasena, M., & Tennekoon, R. (2026). Artificial intelligence and IoT-enabled approaches for COPD detection and monitoring: a systematic literature review. BMC Medical Informatics and Decision Making. https://doi.org/10.1186/s12911-026-03934-y

Image Credits: AI Generated

DOI: 10.1186/s12911-026-03934-y

Keywords: COPD, artificial intelligence, Internet of Things, systematic review, remote monitoring, exacerbation prediction, wearable devices, machine learning, deep learning, digital health, early detection, PRISMA