Performance and Utility of the Sybil Deep Learning Model for Lung Cancer Risk Prediction in Asian High- and Low-Risk Populations
CHEST · 2026-10 · 170 (4) · 1311-1325
DOI: 10.1016/j.chest.2026.05.002 ↗Kurz gefasst
Fachartikel in der Zeitschrift CHEST, 2026-10. Verfasst von Yeon Wook Kim und 8 weiteren Autor:innen. Beteiligte Einrichtung: Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam, South Korea;…. Hauptthemen laut MEDLINE: Lung Neoplasms, Deep Learning, Early Detection of Cancer. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 39 Literaturangaben. Gefördert von: AstraZeneca PLC, Seoul National University Bundang Hospital.
Abstract (Originalsprache)
BACKGROUND: Sybil is a deep learning model designed to predict future lung cancer risk based on a single low-dose chest CT (LDCT) scan, facilitating a precision-based approach to lung cancer screening (LCS). RESEARCH QUESTION: What are the performance and utility of Sybil in a pragmatic Asian LCS cohort? STUDY DESIGN AND METHODS: We analyzed 21,087 individuals 50 to 80 years of age who underwent LDCT screening between January 2009 and December 2021. Baseline LDCT scans were evaluated using Sybil to calculate the risk of lung cancer diagnosis within 1 to 6 years. Model performance was assessed with the area under the receiver operating characteristic curve (AUROC). Stratified analyses were conducted for individuals with ≥ 20 pack-years (PYs) of smoking history (n = 4,611), individuals who ever smoked with < 20 or unknown PYs (n = 5,378), and individuals who have never smoked (INS) (n = 11,098). RESULTS: Among 21,087 participants, subsolid nodules were more common in INS (9.3%) than in those with ≥ 20 PYs (7.1%) or < 20/unknown PYs (6.7%) (P < .001). A total of 257 participants were diagnosed with lung cancer within 6 years. Sybil achieved AUROCs of 0.86 (95% CI, 0.82-0.89) for 1-year invasive lung cancer predictions and 0.74 (95% CI, 0.70-0.77) for 6-year invasive lung cancer predictions. For INS, AUROCs were 0.86 (95% CI, 0.81-0.91) for 1-year predictions and 0.79 (95% CI, 0.74-0.84) for 6-year predictions. Sybil's short-term performance declined in individuals with LDCT findings of granulomatous sequelae, but long-term performance was maintained. Reduced performance was also observed in individuals with subsolid nodules and those without a detected baseline nodule. Five-year overall survival among lung cancer cases was 92.7% in INS, 75.9% for < 20/unknown PYs, and 68.6% in ≥ 20 PYs (P < .001). INTERPRETATION: Our results show that Sybil demonstrated robust performance in predicting future lung cancer in an Asian screening cohort composed of individuals with diverse risk profiles. The findings highlight the potential to develop personalized LCS strategies using Sybil to address region-specific needs.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam, South Korea; Division of Pulmonary and Critical Care Medicine, Department of Inter…
- Seoul National University College of Medicine, Seoul, South Korea; Center for Artificial Intelligence in Healthcare, Seoul National University Bundang Hospital Healthcare Innovation Park, Seongnam, South Korea.
- Center for Artificial Intelligence in Healthcare, Seoul National University Bundang Hospital Healthcare Innovation Park, Seongnam, South Korea.
- Big Data Center, Seoul National University Bundang Hospital, Seongnam, South Korea.
- Department of Radiology, Seoul National University Bundang Hospital, Seongnam, South Korea; Department of Radiology, Seoul National University College of Medicine, Seoul, South Korea.
Förderung
- AstraZeneca PLC
- Seoul National University Bundang Hospital
Zitieren
APA
Kim, Y. W., Oh, J., Park, J., Kim, M., Kim, D. Y., Nam, J. G., Park, J. S., Lee, J. H., & Lee, C. T. (2026). Performance and Utility of the Sybil Deep Learning Model for Lung Cancer Risk Prediction in Asian High- and Low-Risk Populations. CHEST, 170(4), 1311-1325. https://doi.org/10.1016/j.chest.2026.05.002
Vancouver
Kim YW, Oh J, Park J, Kim M, Kim DY, Nam JG, et al. Performance and Utility of the Sybil Deep Learning Model for Lung Cancer Risk Prediction in Asian High- and Low-Risk Populations. CHEST. 2026;170(4):1311-1325. doi:10.1016/j.chest.2026.05.002
BibTeX
@article{Kim2026,
author = {Kim, Yeon Wook and Oh, Jeongbin and Park, Jinyong and Kim, Myeongju and Kim, Dong Yeong and Nam, Ju Gang and Park, Jong Sun and Lee, Jae Ho and Lee, Choon-Taek},
title = {Performance and Utility of the Sybil Deep Learning Model for Lung Cancer Risk Prediction in Asian High- and Low-Risk Populations},
journal = {CHEST},
year = {2026},
volume = {170},
number = {4},
pages = {1311--1325},
doi = {10.1016/j.chest.2026.05.002}
}
EUVIMED