Machine learning algorithm improves detection of NASH (NAS-based) and at-risk NASH, a development and validation study.
Lee J., Westphal M., Vali Y., Boursier J., Ostroff R., Alexander L., Chen Y., Fournier C., Geier A., Francque S., Wonders K., Tiniakos D., Bedossa P., Allison M., Papatheodoridis G., Cortez-Pinto H., Pais R., Dufour J-F., Leeming DJ., Harrison S., Cobbold J., Holleboom AG., Yki-Järvinen H., Crespo J., Ekstedt M., Aithal GP., Bugianesi E., Romero-Gomez M., Karsdal M., Yunis C., Schattenberg JM., Schuppan D., Ratziu V., Brass C., Duffin K., Zwinderman K., Pavlides M., Anstee QM., Bossuyt PM., LITMUS investigators None.
BACKGROUND AIMS: Detecting non-alcoholic steatohepatitis (NASH) remains challenging, while at-risk NASH (steatohepatitis and F≥ 2) tends to progress and is of interest for drug development and clinical application. We developed prediction models by supervised machine learning (ML) techniques, with clinical data and biomarkers to stage and grade non-alcoholic fatty liver disease (NAFLD) patients. APPROACH RESULTS: Learning data were collected in the LITMUS Metacohort (966 biopsy-proven NAFLD adults), staged and graded according to NASH-CRN. Conditions of interest were clinical trial definition of NASH (NAS≥ 4;53%), at-risk NASH (NASH with F≥ 2;35%), significant (F≥ 2;47%) and advanced fibrosis (F≥ 3;28%). Thirty-five predictors were included. Missing data were handled by multiple imputation. Data were randomly split into training/validation (75/25) sets. Gradient boosting machine (GBM) was applied to develop two models for each condition: clinical versus extended (clinical and biomarkers). Two variants of the NASH and at-risk NASH models were constructed: direct and composite models.Clinical GBM models for steatosis/inflammation/ballooning had AUCs of 0.94/0.79/0.72. There were no improvements when biomarkers were included. The direct NASH model produced AUCs (clinical/extended) of 0.61/0.65. The composite NASH model performed significantly better (0.71) for both variants. The composite at-risk NASH model had an AUC of 0.83 (clinical and extended), an improvement over the direct model. Significant fibrosis models had AUCs (clinical/extended) of 0.76/0.78. The extended advanced fibrosis model (0.86) performed significantly better than the clinical version (0.82). CONCLUSIONS: Detection of NASH and at-risk NASH can be improved by constructing independent ML models for each component, using only clinical predictors. Adding biomarkers only improved accuracy for fibrosis.