Joint Research Team from Asan Medical Center and Samsung Medical Center

Quantifying Fibrosis Progression with CT Imaging Analysis

"Risk of Lung Transplant or Death Rises 2.8 Times if Threshold is Exceeded"

Korean researchers have established reference values that quantify the progression of idiopathic pulmonary fibrosis using artificial intelligence (AI)-powered imaging analysis, enabling the prediction of patient outcomes. The study showed that patients whose lung fibrosis score increased by more than 4 percentage points within a year faced a significantly higher risk of either requiring a lung transplant or succumbing to the disease.

(From left) Jooae Choi, Department of Radiology at Asan Medical Center; Hochol Kim, Department of Pulmonology; Hoyun Lee, Department of Radiology at Samsung Seoul Hospital. Asan Medical Center

(From left) Jooae Choi, Department of Radiology at Asan Medical Center; Hochol Kim, Department of Pulmonology; Hoyun Lee, Department of Radiology at Samsung Seoul Hospital. Asan Medical Center

View original image

On August 3, Asan Medical Center, Seoul announced that a team led by Professors Juae Choi (Department of Radiology) and Ho Chul Kim (Department of Pulmonology) at Asan Medical Center, and Professor Ho-yeon Lee (Department of Radiology) at Samsung Medical Center, developed objective criteria to assess disease progression by automatically measuring the extent of lung fibrosis in computed tomography (CT) images using AI software. The results were recently published in the renowned journal 'American Journal of Respiratory and Critical Care Medicine,' which is a leading publication in the fields of respiratory and critical care medicine.


Idiopathic pulmonary fibrosis is a chronic progressive lung disease of unknown cause, characterized by hardening of the lungs and declining respiratory function. Its five-year survival rate after diagnosis is lower than that of major cancers, and each patient’s rate of disease progression varies greatly, making accurate follow-up crucial. Traditional pulmonary function tests only reveal overall lung changes and not the specific location or extent of fibrosis, while CT interpretation can differ between readers, presenting inherent limitations.


The researchers used chest CT images to have the AI automatically calculate the reticulation (net-like patterns) and honeycombing (honeycomb-like patterns), thus quantifying the extent of fibrosis. With this standardized approach, repeated measurements are possible, allowing objective comparison of changes over time. The team analyzed data from 524 patients at Asan Medical Center and conducted external validation using data from 224 patients at Samsung Medical Center.


The analysis found that increases in the fibrosis score were closely linked to worsening lung function. When forced vital capacity decreased by 5%, the fibrosis score increased by 2.72 percentage points; when the diffusing capacity for carbon monoxide dropped by 10%, it rose by 4.52 percentage points.


Based on these findings, the research team proposed that an increase of 4.05 percentage points in the fibrosis score represents a clinically meaningful threshold for disease progression. Patients whose fibrosis score increased beyond this threshold during a year had a significantly higher risk of lung transplantation or death than those who did not, and this risk was approximately 2.8 times higher in the external validation group. When the fibrosis score was incorporated into existing prognostic models, the models’ predictive power improved as well.



Professor Choi explained, “Visual assessment is subject to inter-reader variability and makes it difficult to objectively measure change. Now that we have presented a reference value for clinically meaningful change, this metric can serve as a supplementary index for regular follow-up and treatment strategy planning.” Professor Kim added, “Since idiopathic pulmonary fibrosis is difficult to recover from once the lung is damaged, early diagnosis is crucial. The fact that we confirmed the potential for this to serve as an imaging biomarker in future clinical trials underscores the significance of our findings.”


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing