AI-supported imaging insights platform for COPD care management

Building on a long-lasting collaboration with University Medical Centre Groningen (UMCG) in the field of AI-supported procedures for Lung Volume Reduction (LVR), the SPIDER project (Smart Platform for Integrated Decision-support and Evaluation using Radiology) aims to develop a structured decision-support platform, powered by AI, that transforms CT scans into quantitative visual reports for the assessment of structural lung damage in COPD. At the core of the platform is an intuitive spider web visualization, displaying all clinically relevant imaging metrics in a single, easy-to-interpret graphicThe goal is for SPIDER the resulting platform to support more consistent, data-informed treatment decisions for patients with moderate to severe COPD.  

Combining Thirona’s AI imaging expertise in COPD and UMCG’s clinical leadership in advanced interventional therapies in COPD, the joint initiative aims to develop and validate a platform that provides all clinically relevant imaging insights, using normative reference values. Once integrated into clinical workflow, the platform will support more consistent, personalized, data-informed treatment planning, potentially reducing care costs and ultimately improving clinical outcomes for COPD patients. 

Project highlights

  • Integrated patient reporting, translating quantitive imaging metrics into clinically relevant insights
  • Determination of normative reference values for key COPD imaging metrics
  • Tailored to Multi Disciplinary Teams decision making
  • Validation on real-world data from 1,030 COPD patients

Contribution by Thirona

In this project, Thirona is responsible for developing and validating a functional, visualized, AI-based CT reporting platform that features spider web visualization and integrates all COPD-related longitudinal imaging metrics.

Specifically, this platform will include assessments of emphysema, trapped air, fissure completeness, and other parenchymal measurements; a wide range of bronchial analyses, including quantification of bronchial wall thickening, bronchial dilatation, and bronchial tapering; mucus plug analysis; vascular measurements; and more. 

In parallel, UMCG will use LungQ® to derive normative reference values (Z-scores) from 1,400 CT scans of a non-COPD reference population. These Z-scores will be integrated into the spider web visualization to contextualize individual patients’ results against a non-diseased population. 

Following the final evaluation of the platform on a real-world dataset of 1,030 COPD cases, Thirona will collaborate with UMCG to develop and validate a secure infrastructure for further clinical integration.

Relevant Publications

No relevant publications found.

Learn how LungQ® can help advance treatment of muco-obstructive airway disease