Optimization Through Precision: Imaging-Based Biomarkers and Deep Machine Learning in Clinical Trial Development

1. Understanding Lung Diseases: IPF and ILD

Idiopathic Pulmonary Fibrosis (IPF) is a chronic, progressive scarring disorder of the lungs with no known cause, primarily affecting older adults and commonly carrying a poor prognosis—median survival ranges from 3 to 5 years after diagnosis. Interstitial Lung Diseases (ILDs) encompass a broad group of over 200 disorders, including IPF, autoimmune-related forms (e.g., rheumatoid arthritis–associated ILD), occupational exposures, drug-induced, and hypersensitivity pneumonitis. These conditions are characterized by inflammation and fibrosis of the interstitial lung tissue, disrupting oxygen transfer and leading to symptoms like dyspnea, dry cough, and reduced exercise tolerance—often presenting only after significant lung damage has occurred. Diagnosis typically involves High-Resolution Computed Tomography (HRCT) scans, pulmonary function tests, and sometimes biopsies. Given their subtle onset and varied progression, ILDs, especially IPF, are challenging to monitor making them prime candidates for AI-powered imaging tools that enable earlier detection and better therapeutic evaluation ^1–4.

2. Why Traditional Imaging Falls Short

Human analysis of lung images is often limited by subjectivity and variability between readers ⁵. AI addresses these issues by offering consistent, objective interpretation across large imaging datasets. AI tools can detect minute structural changes invisible to the human eye, track fibrosis progression over time, and provide quantitative assessments. These capabilities are particularly relevant in diseases such as IPF, where early and accurate monitoring may significantly impact outcomes ⁴.

3. AI-Powered Lung Imaging: What It Is and How It Works

AI-powered lung imaging refers to the application of deep learning algorithms to CT scans to detect early-stage fibrosis, airway remodeling, and other disease markers. This technology is already being integrated into respiratory medicine workflows to support clinical trial design and patient stratification ⁶.

4. Fortrea’s Role in Enabling Innovation

Fortrea contributes to the integration of AI tools in clinical research by identifying and onboarding innovative vendors to enhance the scientific and operational quality of its clinical trials ⁷. Our vendor management approach helps in seamless integration of cutting-edge tools while maintaining regulatory compliance and operational efficiency. From early evaluation and feasibility assessments to long-term collaboration and data oversight, Fortrea prioritizes vendor collaboration that aligns with our mission to deliver patient-centric, technology-enabled clinical solutions. This strategic engagement model supports accelerated timelines, robust data collection, and innovative trial designs across respiratory and other therapeutic areas.

Fortrea has positioned itself at the forefront of AI integration in respiratory trials. Through strategic vendor collaboration and internal platforms like the Xcellerate Ecosystem, Fortrea ensures that AI tools are not only scientifically robust but also operationally scalable. The company’s AI Innovation Studio supports real-time data curation, risk monitoring, and predictive analytics—streamlining trial delivery while enhancing patient safety and data quality.

5. A Tech-Bio Innovator in Imaging Analytics

Qureight is a UK-based company using deep learning to analyze high-resolution chest CT (HRCT) scans in interstitial lung disease, pulmonary hypertension, and lung cancer ⁶. Leveraging its imaging biomarkers and large clinical datasets, Qureight also generates synthetic control arms that model patient response trajectories, reducing trial duration and limiting the need for placebo groups ⁸. Endeavor Biomedicines used Qureight’s imaging biomarkers in a 12-week Phase IIa IPF study of 36 patients to demonstrate greater effect sizes than FVC, with clear associations between treatment response and changes in lung volume and pulmonary vascular volume ⁹.

Qureight’s HRCT analysis has been successfully deployed in multiple Phase II and III IPF trials, where it enables rapid image evaluation, streamlines remote site qualification, and delivers reproducible quantitative endpoints—capabilities that are critical in complex, multicenter trials ^9,10. The company is increasingly recognized as a scientific collaborator for clinical development in respiratory disease, offering both regulatory-grade imaging science and scalable technical integration.

6. Benefits for Patients

AI-powered lung imaging offers patients, outside of clinical trials, the potential for earlier diagnosis, which can lead to timely interventions and improved outcomes. Additionally, Qureight’s platform accepts imaging from all HRCT machines, and its simple upload process allows sites in complex studies to share data quickly. This capability accelerates patient qualifications and reduces study start-up times, helping patients access innovative therapies sooner.

7. Conclusion

AI and machine learning tools are increasingly being used to support respiratory clinical trials. By reducing site burden, enhancing data quality, and accelerating trial timelines, these technologies are advancing both operational efficiency and scientific rigor. Fortrea’s engagement in implementing such tools demonstrates its commitment to innovation and patient-centric trial design. As Fortrea continues to invest in AI-driven innovation, the future of lung imaging looks smarter—powered by data and guided by science.

References

Richeldi, L., Collard, H. R., & Jones, M. G. (2017). Idiopathic pulmonary fibrosis. Lancet, 389(10082), 1941–1952.
Ryerson, C. J., et al. (2025). Standardized Clinical Terms and Definitions for Interstitial Lung Disease: A Consensus Statement from the Fleischner Society. American Journal of Respiratory and Critical Care Medicine.
Walsh, S. L. F. (2018). Imaging biomarkers and staging in IPF. Current Opinion in Pulmonary Medicine, 24(5), 445–452.
Walsh, S. L. F., et al. (2020). Imaging research in fibrotic lung disease: applying deep learning to unsolved problems. Lancet Respiratory Medicine.
Walsh, S. L., et al. (2015). Interobserver agreement for the ATS/ERS/JRS/ALAT criteria for a UIP pattern on CT. Thorax.
Thillai, M., et al. (2024). Deep learning-based segmentation of CT scans predicts disease progression and mortality in IPF. American Journal of Respiratory and Critical Care Medicine.
Fortrea launches AI Innovation Studio to galvanize technology and human solutions to improve clinical trial delivery. (2024). GlobeNewswire.
Kirov, K. R., B. E., Thillai, M., Woodhead, F. A., Lazarus, H. M., Conoscenti, C. S., & Walsh, S. L. F. (2025). Monte Carlo External Control Arm Generation Utilising Real-world Patient Data and Deep Learning-based Quantitative CT Metrics Demonstrates Treatment Effect in the Atlas IPF Trial. In American Thoracic Society Congress, American Thoracic Society, San Francisco, USA, A2090–A2090.
Walsh, S. L. F., D. A., Hood, J., de los Rios, M., Montiero, M., Bussell, E., & Thillai, M. (2025). Deep learning-based disease severity biomarkers on CT: Posthoc analysis in a Phase 2a placebo-controlled study of ENV-101 in subjects with idiopathic pulmonary fibrosis. In American Thoracic Society Conference. American Thoracic Society.