AI capable of predicting COPD exacerbations with continuous temperature data
Click Here to Manage Email Alerts
Key takeaways:
- An AI had high performance in predicting COPD exacerbations using patients’ body temperature data.
- Thirty-seven patients wore an armband that continuously monitored temperature for a total of 970 weeks.
Using continuous temperature monitoring data from patients with COPD, an AI was able to predict exacerbations, according to a poster presented at the European Respiratory Society International Congress.
“Continuous temperature monitoring is feasible in patients with COPD and might aid the early detection of exacerbations,” Mairi MacLeod, MD, clinical research fellow at the National Heart and Lung Institute, Imperial College London, and colleagues wrote on the poster.
In this feasibility study, MacLeod and colleagues evaluated 37 patients (mean age, 71.8 years; 51% men) with COPD from the London COPD exacerbation (EXCEL) cohort to determine if continuous at-home temperature monitoring via an armband sensor helps identify exacerbations early.
Researchers uncovered the value of temperature monitoring in relation to COPD exacerbations through an AI trained with previous exacerbations. According to the poster, they assessed how well the AI could predict an exacerbation based on temperature signature data from the past 7 days.
During a monitoring period of 970 weeks, researchers found that 30 exacerbations — made known by patient reports and daily symptom diary cards — took place. The poster indicated that five exacerbations could not be included in the AI assessment due to insufficient temperature trace data quality.
The AI demonstrated 84% sensitivity (predicted 21 out of 25 exacerbations) and 90.6% specificity, according to the poster.
“Future work in larger cohorts should be undertaken to confirm this and the possible benefits of early detection and intervention in this group,” MacLeod and colleagues wrote on the poster.
Perspective
Back to Top
Alanna Kavanaugh, EdD(c), FNP-BC, CCRN, FCCP, FFNMRCSI
The integration of AI into medical practice has sparked considerable debate, but recent findings on AI predictive modeling for COPD exacerbations suggest we may be on the cusp of a significant breakthrough in chronic disease management.
In pulmonary medicine, AI is increasingly used to enhance disease detection, monitoring and prognosis. Computer-aided diagnostic tools, for example, have proven effective in identifying various lung diseases, particularly in image-based specialties such as radiology.
AI also shows promise in managing interstitial lung diseases, where it analyzes CT scans and pulmonary function tests to predict disease progression. However, gaps remain, and more research is needed to refine these tools for holistic ILD management.
By applying similar AI-driven models to COPD exacerbations, as demonstrated in a recent study that used wearable temperature sensors and daily symptom checklists, health care providers could intervene earlier to prevent hospitalizations.
The study showed impressive results: AI predicted COPD exacerbations with 84% sensitivity and 90.6% specificity, based on temperature changes over 7 days. Of 25 diagnosed exacerbations, the AI successfully predicted 21, while producing only three false positives across 32 control regions.
These findings are both surprising and significant. The ability to predict exacerbations before they fully manifest could profoundly impact patient care, shifting the model from reactive to proactive treatment.
Given that COPD exacerbations are associated with increased hospitalizations, reduced quality of life and higher health care costs, early intervention via AI-driven alerts could be a game-changer. As a clinician, I’ve seen firsthand how timely treatment can improve outcomes and reduce costs.
However, this study has limitations, including its small sample size of 37 patients over 970 weeks. More research with diverse populations is needed to validate these findings. Additionally, integrating AI alerts into clinical workflows and ensuring patient and provider trust are critical challenges. Will they rely on these systems enough to change their care practices?
Moving forward, research should focus on real-world applications, examining AI’s impact on reducing hospitalizations, improving quality of life and cost-effectiveness. Incorporating additional clinical data, such as oxygen levels and heart rate, could enhance predictive accuracy further.
While AI in medicine remains a topic of debate, studies like this demonstrate its potential to revolutionize chronic disease management. With continued research, AI could become a powerful tool in managing COPD and beyond.
References:
Dack E, et al. Invest Radiol. 2023;doi:10.1097/RLI.0000000000000974.
Mannino, DM, et al. Lancet. 2007;doi:10.1016/S0140-6736(07)61380-4.
Topol EJ, Nat Med. 2019;doi:10.1038/s41591-018-0300-7.
Collapse
MacLeod M, et al. Detecting COPD exacerbations using AI predictive modelling with continuous temperature sensing. Presented at: European Respiratory Society International Congress; Sept. 7-11, 2024; Vienna.
Read more about
Click Here to Manage Email Alerts