APSR2026-775

Yes
Late-Breaking
Submitted
Poster Presentation
 
AP10-025
P25
 
Format: PPTX Only
https://storage.unitedwebnetwork.com/files/1285/1221060-46904-APSR-AP10-025_440125.pdf
Format: PDF Only
07. COPD
13

No more than 15 authors can be listed.
Please ensure the authors are listed in the right order.
If there are multiple affiliations, please use the "+" button next to "Select Country/Region" to add additional affiliation fields.

Vipul Kumar Garg * Yashoda Hospitals , Hitech city Pulmonary Medicine Hyderabad India
Pranitha Peddakurma - Yashoda Hospitals , Hitech city Pulmonary Medicine Hyderabad India
Rinoosha Rachel - Yashoda Hospitals , Hitech city Pulmonary Medicine Hyderabad India
Jonnalagadda Aswith Chowdary - Yashoda Hospitals , Hitech city Pulmonary Medicine Hyderabad India
Virender Pratibh Prasad - Yashoda Hospitals , Hitech city Pulmonary Medicine Hyderabad India
Venkata Nagarjuna Maturu - Yashoda Hospitals , Hitech city Pulmonary Medicine Hyderabad India
Pravalika Nanavathula - Yashoda Hospitals , Hitech city Pulmonary Function Laboratory Hyderabad India
Mohammed Ashrafuddin Tameez - Yashoda Hospitals , Hitech city Pulmonary Function Laboratory Hyderabad India
Shivani Acharya - Qure ai Evidence Generation Mumbai India
Mir Siptain Ali - Qure ai Evidence Generation Mumbai India
Parasaran - Qure ai Evidence Generation Mumbai India
Rejesh Bose - Qure ai Evidence Generation Mumbai India
Rohitashva Agrawal - Qure ai Evidence Generation Mumbai India
 
 
Presenting Author Details
Vipul Kumar
Garg
gargvipul89@gmail.com
+919949779680
31-40
India
Original Research
Real-World Clinical Utility of AI-Assisted Chest X-ray interpretation for diagnosing COPD
Maximum 25 words
Abstract Content
Background: Chronic Obstructive Pulmonary Disease (COPD) remains substantially underdiagnosed in low-resource settings, largely due to limited access to spirometry and specialist care. Chest radiography is widely performed across clinical settings and offers a potential screening opportunity. Artificial intelligence (AI)-based analysis of chest radiographs may enable clinician-directed triage by identifying patients at high risk for COPD who may benefit from confirmatory testing. Aims: To evaluate the real-world clinical utility of an AI-assisted chest X-ray pathway for COPD detection in routine practice.
This study included data from a tertiary care center between September 2025 and February 2026. A total of 15,213 consecutive chest X-rays were analyzed using an AI model, of which 332 patients had paired spirometry data. COPD was defined according to Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria as a post-bronchodilator FEV₁/FVC ratio <0.70 (with additional analysis using LLN, where applicable). Predictive performance (PPV and NPV) was assessed in patients with paired AI and spirometry results.
Among 332 patients with paired AI and spirometry data, 165 were classified as high risk for COPD, of whom 113 had spirometry-confirmed COPD (PPV: 68.5%). Among 167 classified as low risk, 119 had normal spirometry (NPV: 71.3%). In the health check-up subgroup analysis (n=137), 37 of 66 patients classified as high risk had spirometry-confirmed COPD (PPV: 56.1%), while 59 of 71 classified as low risk had normal spirometry (NPV: 83.1%), indicating improved rule-out performance in this lower-prevalence, asymptomatic population.
AI-assisted analysis of routine chest radiographs demonstrated moderate predictive performancefor COPD detection in real-world practice. This approach may facilitate opportunistic case-finding and efficient prioritization for confirmatory spirometry, particularly in resource-constrained settings.
 
 
 
 
 
No
No
 
JPG or JPEG or PNG format ONLY
Chronic Obstructive Pulmonary Disease, Artificial Intelligence, Chest X-ray,
maximum 3 keywords and no abbreviations, separated by commas
Some co-authors are affiliated with Qure.ai, the developers of the AI model used. This study represents a non-financial academic collaboration with no associated commercial or financial interests and was conducted with Institutional Ethics Committee approval.
277