Dataset Information

Machine learning-based clinical decision support system for treatment recommendation and overall survival prediction of hepatocellular carcinoma: a multi-center study.

ABSTRACT: The treatment decisions for patients with hepatocellular carcinoma are determined by a wide range of factors, and there is a significant difference between the recommendations of widely used staging systems and the actual initial treatment choices. Herein, we propose a machine learning-based clinical decision support system suitable for use in multi-center settings. We collected data from nine institutions in South Korea for training and validation datasets. The internal and external datasets included 935 and 1750 patients, respectively. We developed a model with 20 clinical variables consisting of two stages: the first stage which recommends initial treatment using an ensemble voting machine, and the second stage, which predicts post-treatment survival using a random survival forest algorithm. We derived the first and second treatment options from the results with the highest and the second-highest probabilities given by the ensemble model and predicted their post-treatment survival. When only the first treatment option was accepted, the mean accuracy of treatment recommendation in the internal and external datasets was 67.27% and 55.34%, respectively. The accuracy increased to 87.27% and 86.06%, respectively, when the second option was included as the correct answer. Harrell's C index, integrated time-dependent AUC curve, and integrated Brier score of survival prediction in the internal and external datasets were 0.8381 and 0.7767, 91.89 and 86.48, 0.12, and 0.14, respectively. The proposed system can assist physicians by providing data-driven predictions for reference from other larger institutions or other physicians within the same institution when making treatment decisions.

SUBMITTER: Lee KH

PROVIDER: S-EPMC10770025 | biostudies-literature | 2024 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning-based clinical decision support system for treatment recommendation and overall survival prediction of hepatocellular carcinoma: a multi-center study.

Lee Kyung Hwa KH Choi Gwang Hyeon GH Yun Jihye J Choi Jonggi J Goh Myung Ji MJ Sinn Dong Hyun DH Jin Young Joo YJ Kim Minseok Albert MA Yu Su Jong SJ Jang Sangmi S Lee Soon Kyu SK Jang Jeong Won JW Lee Jae Seung JS Kim Do Young DY Cho Young Youn YY Kim Hyung Joon HJ Kim Sehwa S Kim Ji Hoon JH Kim Namkug N Kim Kang Mo KM

NPJ digital medicine 20240105 1

The treatment decisions for patients with hepatocellular carcinoma are determined by a wide range of factors, and there is a significant difference between the recommendations of widely used staging systems and the actual initial treatment choices. Herein, we propose a machine learning-based clinical decision support system suitable for use in multi-center settings. We collected data from nine institutions in South Korea for training and validation datasets. The internal and external datasets in ...[more]

PMID: 38182886

Similar Datasets

Project description:Background and Objectives: Hepatocellular carcinoma (HCC) is the most common form of primary liver cancer with a highly unfavorable prognosis. Aims: Retrospective statistical analysis of patients with HCC in the field of liver cirrhosis treated at our center from the perspective of demography, and the effects of key changes in diagnostic and therapeutic procedures in the last 10 years on overall survival (OS) and earlier diagnosis. Materials and Methods: This study included 170 cirrhotic patients with HCC (136 men, 80%). Demographic and etiological factors and OS were analyzed based on distribution into three groups according to the period and key changes in diagnostic and therapeutic approaches (BCLC classification staging; standardization of protocol for transarterial chemoembolization (TACE) and the introduction of direct-acting antivirals (DAA) for the treatment of chronic viral hepatitis C (HCV); expansion of systemic oncological therapy). Results: The mean age at the time of diagnosis was 69.3 years (SD = 8.1), and etiology was as follows: non-alcoholic steatohepatitis (NASH) 39%, alcoholic liver disease (ALD) 36%, HCV 18%, cryptogenic liver cirrhosis 3%, chronic hepatitis B infection (HBV) 2%, and other etiology 2%. Distribution of stages according to the BCLC: 0 + A 36%, B 31%, C 22%, and D 11%. However, the distribution in the first studied period was as follows: 0 + A 15%, B 34%, C 36%, and D 15%; and in the last period: 0 + A 45%, B 27%, C 17%, and D 11%, and difference was statistically significant (p < 0.05). The median OS for stages 0 + A, B, C, and D was 58, 19, 6, and 2 months, respectively. During the monitored period, there was a visible increase in the etiology of ALD from 30% to 47% and a decrease in HCV from 22% to 11%. In patients treated with TACE (stage B), the median OS grew from 10 to 24 months (p < 0.0001) between the marginal monitored periods. Conclusions: We described a decreasing number of patients with HCV-related HCC during follow-up possibly linked with the introduction of DAA. In our cohort, an improvement in early-stage diagnosis was found, which we mainly concluded as a result of proper ultrasound surveillance, the institution of a HCV treatment center, and increased experience of our sonographers with an examination of cirrhotic patients. Lastly, we described significantly improved overall survival in patients with intermediate HCC treated by TACE, due to the increased experience of interventional radiologists with the method at our facility and an earlier switch to systemic therapy in case of non-response to TACE.

Dataset Information

Machine learning-based clinical decision support system for treatment recommendation and overall survival prediction of hepatocellular carcinoma: a multi-center study.

Publications

Machine learning-based clinical decision support system for treatment recommendation and overall survival prediction of hepatocellular carcinoma: a multi-center study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets