Machine learning predicts liver cancer risk from routine clinical data: a large population-based multicentric study
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background and aims
Hepatocellular carcinoma (HCC) is a highly fatal tumor, for which early detection and risk stratification is crucial, yet remains challenging. We aimed to develop an interpretable machine-learning framework for HCC risk stratification based on routinely collected clinical data.
Methods
We leverage data obtained from over 900,000 individuals and 983 cases of HCC across two large-scale population-based cohorts: the UK Biobank study and the “All Of Us Research Program”. For all of these patients, clinical data from timepoints years before diagnosis of HCC was available. We integrate data modalities including demographics, electronic health records, lifestyle, routine blood tests, genomics and metabolomics to offer a unique, multi-modal perspective on HCC risk.
Results
Our random-forest-based model significantly outperforms all publicly available state-of-the-art risk-scores, with an AUROC of 0.88 both for internal and external test sets. We demonstrate robustness of our model across ethnic subgroups, a major advance over previous models with variable performance by ethnicity. Further, we perform extensive feature-importance analysis, showcasing our approach as an interpretable framework. We provide all model weights and an open-source web calculator to facili-tate further validation of our model.
Conclusion
Our study presents a robust and interpretable machine-learning framework for HCC risk stratification, which offers the potential to improve early detection and could ultimately reduce disease burden through targeted interventions.
Lay summary
Finding liver cancer early is crucial for successful treatment. Therefore, screening with abdominal ultra-sound can be performed. However, it is not clear who should receive ultrasound screening, as with the current standard of screening only patients with liver cirrhosis, a severe liver disease, many patients are diagnosed with liver cancer in late stages. Therefore, we trained a machine learning model, acting like many decision trees at the same time, to detect patients with high risk of liver cancer by looking at patterns of almost 1000 cases of liver cancer in a population of 900.000 individuals. In a separate set of patients, which the model has not seen during training, our model worked better than all available models. Additionally, we investigated 1. how the model comes to its prediction, 2. whether it works in males and females alike and 3. which data is most relevant for the model. Like this, our model can help sort patients into categories like “high-risk”, “medium-risk” and “low-risk”, via which screening strategies can then be decided, to help improve early detection of liver cancer.
