EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Machine Learning Prediction of Liver Fibrosis in Patients With Metabolic Dysfunction–Associated Steatotic Liver Disease

Salvatore Petta, Grazia Pennisi, Ciro Celsa, Sofiane Messaoudi, Emmanuel Tsochatzis, Elisabetta Bugianesi, Masato Yoneda, Ming-Hua Zheng, Hannes Hagström, Jérôme Boursier, José Luis Calleja, George Boon-Bee Goh und 35 weitere

Clinical Gastroenterology and Hepatology · 2026-07

DOI: 10.1016/j.cgh.2026.07.006 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Clinical Gastroenterology and Hepatology, 2026-07. Verfasst von Salvatore Petta und 46 weiteren Autor:innen. Beteiligte Einrichtung: Section of Gastroenterology and Hepatology, Department of Health Promotion, Mother and Child Care, Internal Medicine, and Medical Specialties (PROMIS…. Die Arbeit nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Health, Ministero dell'Istruzione dell'Universita e della Ricerca.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

BACKGROUND & AIMS: Accurate staging of liver fibrosis is crucial for risk stratification in patients with metabolic dysfunction-associated steatotic liver disease. We aimed to develop and validate artificial intelligence-based models capable of distinguishing fibrosis stages. METHODS: We developed and validated machine learning models to predict fibrosis stages in more than 3600 biopsy-confirmed patients with metabolic dysfunction-associated steatotic liver disease using 22 clinical features, liver stiffness measurement, and controlled attenuation parameter. Models included Feature Tokenizer Transformer, TabNet variants with distance-aware losses, ordinal Multilayer Perceptron with the COnditional RAnk Logits framework. Three centers were prespecified for geographic external validation. In the remaining centers, we used a stratified 75/25 split into a training pool and an internal random test set, tuned hyperparameters by stratified 10-fold cross-validation on the training pool, and trained 1 model per imputed dataset (M = 5) with an internal 80/20 train-validation split for early stopping and operating-point selection. Performance was pooled across imputations using Rubin's rules and reported for the internal random test set and the held-out centers. RESULTS: Feature Tokenizer Transformer and TabNet achieved the highest performance in binary tasks: area under the receiver operating characteristic curve = 0.860 and 0.855 (F ≥3 vs 0-F2) and area under the receiver operating characteristic curve = 0.800 and 0.788 (F ≥2 vs 0-F1), significantly outperforming the Fibrosis-4 index. For F ≥3, Feature Tokenizer Transformer had significantly higher area under the receiver operating characteristic curve than liver stiffness measurement (P = .014) but only marginally higher than AGILE3+, and, together with TabNet, the highest accuracy and the lowest gray zone (8.2% and 8.4%, respectively). For multiclass staging, ordinal models performed best with Multilayer Perceptron with the Consistent Rank Logits achieving quadratic weighted kappa = 0.616. CONCLUSIONS: Deep learning models with ordinal-aware architectures can accurately predict liver fibrosis stages using routinely available clinical data, offering a scalable alternative to biopsy without requiring specialized biomarkers.

Schlagwörter

Stichwörter der Autor:innen

Deep LearningLiver FibrosisMASLDMachine LearningNoninvasive DiagnosisOrdinal Classification

Autor:innen und Institutionen

  • Section of Gastroenterology and Hepatology, Department of Health Promotion, Mother and Child Care, Internal Medicine, and Medical Specialties (PROMISE), University of Palermo, Palermo, Italy.
  • Section of Gastroenterology and Hepatology, Department of Health Promotion, Mother and Child Care, Internal Medicine, and Medical Specialties (PROMISE), University of Palermo, Palermo, Italy. Electronic address: grazia.…
  • Department of Engineering, University of Palermo, Palermo, Italy.
  • University College London Institute for Liver and Digestive Health, Royal Free Hospital and University College London, London, United Kingdom.
  • Division of Gastroenterology, Department of Medical Sciences, A.O.U. Città della Salute e della Scienza di Torino, University of Turin, Turin, Italy.
  • Department of Gastroenterology and Hepatology, Yokohama City University Graduate School of Medicine, Yokohama, Japan.

Förderung

  • Ministry of Health
  • Ministero dell'Istruzione dell'Universita e della Ricerca

Zitieren

APA

Petta, S., Pennisi, G., Celsa, C., Messaoudi, S., Tsochatzis, E., Bugianesi, E., Yoneda, M., Zheng, M. H., Hagström, H., Boursier, J., Calleja, J. L., Boon-Bee Goh, G., Chan, W. K., Gallego-Durán, R., Sanyal, A. J., de Lédinghen, V., Newsome, P. N., Fan, J. G., Castéra, L., …, & Cammà, C. (2026). Machine Learning Prediction of Liver Fibrosis in Patients With Metabolic Dysfunction–Associated Steatotic Liver Disease. Clinical Gastroenterology and Hepatology. https://doi.org/10.1016/j.cgh.2026.07.006

Vancouver

Petta S, Pennisi G, Celsa C, Messaoudi S, Tsochatzis E, Bugianesi E, et al. Machine Learning Prediction of Liver Fibrosis in Patients With Metabolic Dysfunction–Associated Steatotic Liver Disease. Clinical Gastroenterology and Hepatology. 2026. doi:10.1016/j.cgh.2026.07.006

BibTeX
@article{Petta2026,
  author = {Petta, Salvatore and Pennisi, Grazia and Celsa, Ciro and Messaoudi, Sofiane and Tsochatzis, Emmanuel and Bugianesi, Elisabetta and Yoneda, Masato and Zheng, Ming-Hua and Hagström, Hannes and Boursier, Jérôme and Calleja, José Luis and Boon-Bee Goh, George and Chan, Wah-Kheong and Gallego-Durán, Rocio and Sanyal, Arun J and de Lédinghen, Victor and Newsome, Philip N and Fan, Jian-Gao and Castéra, Laurent and Lai, Michelle and Fournier-Poizat, Céline and Lai-Hung Wong, Grace and Armandi, Angelo and Nakajima, Atsushi and Liu, Wen-Yue and Shang, Ying and de Saint-Loup, Marc and Llop, Elba and Jun Teh, Kevin Kim and Lara-Romero, Carmen and Asgharpour, Amon and Sau-Wai Chan, Mandy and Romero-Gomez, Manuel and Lin, Huapeng and Kim, Seung Up and Cheuk-Fung Yip, Terry and Calvaruso, Vincenza and Zoncapè, Mirko and Che-To Lai, Jimmy and Yang, Boyu and Lee, Hye Won and Di Maria, Gabriele and Enea, Marco and Contino, Salvatore and Wai-Sun Wong, Vincent and Cirrincione, Giansalvo and Cammà, Calogero},
  title = {Machine Learning Prediction of Liver Fibrosis in Patients With Metabolic Dysfunction–Associated Steatotic Liver Disease},
  journal = {Clinical Gastroenterology and Hepatology},
  year = {2026},
  doi = {10.1016/j.cgh.2026.07.006}
}
Kostenlos starten →