Prediction of the risk of developing hepatocellular carcinoma in health screening examinees: a Korean cohort study

Abstract

Background Almost all Koreans are covered by mandatory national health insurance and are required to undergo health screening at least once every 2 years. We aimed to develop a machine learning model to predict the risk of developing hepatocellular carcinoma (HCC) based on the screening results and insurance claim data. Methods The National Health Insurance Service-National Health Screening database was used for this study (NHIS-2020-2-146). Our study cohort consisted of 417,346 health screening examinees between 2004 and 2007 without cancer history, which was split into training and test cohorts by the examination date, before or after 2005. Robust predictors were selected using Cox proportional hazard regression with 1000 different bootstrapped datasets. Random forest and extreme gradient boosting algorithms were used to develop a prediction model for the 9-year risk of HCC development after screening. After optimizing a prediction model via cross validation in the training cohort, the model was validated in the test cohort. Results Of the total examinees, 0.5% (1799/331,694) and 0.4% (390/85,652) in the training cohort and the test cohort were diagnosed with HCC, respectively. Of the selected predictors, older age, male sex, obesity, abnormal liver function tests, the family history of chronic liver disease, and underlying chronic liver disease, chronic hepatitis virus or human immunodeficiency virus infection, and diabetes mellitus were associated with increased risk, whereas higher income, elevated total cholesterol, and underlying dyslipidemia or schizophrenic/delusional disorders were associated with decreased risk of HCC development (p < 0.001). In the test, our model showed good discrimination and calibration. The C-index, AUC, and Brier skill score were 0.857, 0.873, and 0.078, respectively. Conclusions Machine learning-based model could be used to predict the risk of HCC development based on the health screening examination results and claim data.

The Impact of Artificial Intelligence on Health Equity in Oncology: Scoping Review

“…Several articles in our review applied AI to shed light on the influence of race and socioeconomic status on health outcomes in oncology. For example, An et al [ 143 ] used an ML algorithm to examine the risk factors for the development of hepatocellular carcinoma in a Korean cohort, noting that higher income is associated with a lower risk of developing hepatocellular carcinoma. Bibault et al [ 144 ] applied AI to satellite imagery to investigate the relationship between socioeconomic status and cancer prevalence, observing that “satellite features are highly correlated with individual socioeconomic and health measures that are linked to cancer prevalence.” Several studies have suggested that applying AI to demographic data could help provide more comprehensive risk stratification models in oncology [ 112 , 168 , 169 ].…”

“…For most studies in our review, there was a lack of justification for the use of AI and, more specifically, a lack of discussion as to why particular AI algorithms were chosen and their advantages over other statistical methods to address a given research question. AI algorithms are undeniably powerful tools for analyzing large amounts of data and selecting articles that mention the benefits of AI over other statistical methods [ 143 , 144 , 167 , 168 ]. However, others have argued that the use of AI has not yielded better risk prediction models compared with traditional statistical methods [ 169 ].…”

“…These data are often skewed toward the populations, diseases, and treatments available in countries training and developing AI technologies, thereby decreasing their generalizability to populations in LMICs. Articles from our review addressed this issue, voicing concerns about the applicability of AI algorithms developed in HICs to LMICs [108,110,134,143,[149][150][151].…”

Type 2 Diabetes Mellitus- and Complication-Related Risk of Nontuberculous Mycobacterial Disease in a South Korean Cohort

“…Although we adjusted all comorbid diseases in the multivariate analysis, there exists a possibility of further NTM detection in T2DM patients with a higher number of comorbidities. Finally, because the data concerning human immunodeficiency virus (HIV) are masked in the NHIS-NSC database, as this diagnostic code constitutes sensitive personal information ( 42 ), we were unable to adjust for HIV in our analysis. However, considering that the burden of HIV infection is low in South Korea (incident HIV infection rate in 2011 was 1.8 per 100,000 individuals [ 43 ]) and the prevalence of NTM disease was only approximately 2% (19/1,060) in HIV-infected patients in Southeast Asia ( 44 ), the cases of NTM disease associated with HIV infection would be low in our cohort.…”

Development of Machine Learning-Based Personalized Predictive Models for Early Detection of Hepatocellular Carcinoma in HBV-Related Cirrhosis Patients with Low Levels of Serum Alpha-Fetoprotein

“…(7). For HCC prediction, various deep learning and machine learning algorithms have been employed (8)(9)(10)(11)(12)(13)(14). For instance, Hashem et al constructed several HCC classi cation models based on machine learning algorithms using simple factors such as age, AFP, alkaline phosphate (ALP), albumin, and total bilirubin (15).…”