Predicting hepatocellular carcinoma survival with artificial intelligence
Abstract Despite the extensive research on hepatocellular carcinoma (HCC) exploring various treatment strategies, the survival outcomes have remained unsatisfactory. The aim of this research was to evaluate the ability of machine learning (ML) methods in predicting the survival probability of HCC pa...
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-02-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-90884-6 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849715643356545024 |
|---|---|
| author | İsmet Seven Doğan Bayram Hilal Arslan Fahriye Tuğba Köş Kübranur Gümüşlü Selin Aktürk Esen Mücella Şahin Mehmet Ali Nahit Şendur Doğan Uncu |
| author_facet | İsmet Seven Doğan Bayram Hilal Arslan Fahriye Tuğba Köş Kübranur Gümüşlü Selin Aktürk Esen Mücella Şahin Mehmet Ali Nahit Şendur Doğan Uncu |
| author_sort | İsmet Seven |
| collection | DOAJ |
| description | Abstract Despite the extensive research on hepatocellular carcinoma (HCC) exploring various treatment strategies, the survival outcomes have remained unsatisfactory. The aim of this research was to evaluate the ability of machine learning (ML) methods in predicting the survival probability of HCC patients. The study retrospectively analyzed cases of patients with stage 1–4 HCC. Demographic, clinical, pathological, and laboratory data served as input variables. The researchers employed various feature selection techniques to identify the key predictors of patient mortality. Additionally, the study utilized a range of machine learning methods to model patient survival rates. The study included 393 individuals with HCC. For early-stage patients (stages 1–2), the models reached recall values of up to 91% for 6-month survival prediction. For advanced-stage patients (stage 4), the models achieved accuracy values of up to 92% for 3-year overall survival prediction. To predict whether patients are ex or not, the accuracy was 87.5% when using all 28 features without feature selection with the best performance coming from the implementation of weighted KNN. Further improvements in accuracy, reaching 87.8%, were achieved by applying feature selection methods and using a medium Gaussian SVM. This study demonstrates that machine learning techniques can reliably predict survival probabilities for HCC patients across all disease stages. The research also shows that AI models can accurately identify a high proportion of surviving individuals when assessing various clinical and pathological factors. |
| format | Article |
| id | doaj-art-e966115018ce41078400c45a3ab83bd8 |
| institution | DOAJ |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-e966115018ce41078400c45a3ab83bd82025-08-20T03:13:17ZengNature PortfolioScientific Reports2045-23222025-02-0115111410.1038/s41598-025-90884-6Predicting hepatocellular carcinoma survival with artificial intelligenceİsmet Seven0Doğan Bayram1Hilal Arslan2Fahriye Tuğba Köş3Kübranur Gümüşlü4Selin Aktürk Esen5Mücella Şahin6Mehmet Ali Nahit Şendur7Doğan Uncu8Ankara Bilkent City Hospital, Medical Oncology ClinicAnkara Bilkent City Hospital, Medical Oncology ClinicComputer Engineering Department, Ankara Yıldırım Beyazıt UniversityAnkara Bilkent City Hospital, Medical Oncology ClinicComputer Engineering Department, Ankara Yıldırım Beyazıt UniversityAnkara Bilkent City Hospital, Medical Oncology ClinicDepartment of Internal Medicine, Ankara Bilkent City HospitalAnkara Bilkent City Hospital, Medical Oncology ClinicAnkara Bilkent City Hospital, Medical Oncology ClinicAbstract Despite the extensive research on hepatocellular carcinoma (HCC) exploring various treatment strategies, the survival outcomes have remained unsatisfactory. The aim of this research was to evaluate the ability of machine learning (ML) methods in predicting the survival probability of HCC patients. The study retrospectively analyzed cases of patients with stage 1–4 HCC. Demographic, clinical, pathological, and laboratory data served as input variables. The researchers employed various feature selection techniques to identify the key predictors of patient mortality. Additionally, the study utilized a range of machine learning methods to model patient survival rates. The study included 393 individuals with HCC. For early-stage patients (stages 1–2), the models reached recall values of up to 91% for 6-month survival prediction. For advanced-stage patients (stage 4), the models achieved accuracy values of up to 92% for 3-year overall survival prediction. To predict whether patients are ex or not, the accuracy was 87.5% when using all 28 features without feature selection with the best performance coming from the implementation of weighted KNN. Further improvements in accuracy, reaching 87.8%, were achieved by applying feature selection methods and using a medium Gaussian SVM. This study demonstrates that machine learning techniques can reliably predict survival probabilities for HCC patients across all disease stages. The research also shows that AI models can accurately identify a high proportion of surviving individuals when assessing various clinical and pathological factors.https://doi.org/10.1038/s41598-025-90884-6Artificial intelligenceHepatocellular carcinomaMachine learningSurvival prediction |
| spellingShingle | İsmet Seven Doğan Bayram Hilal Arslan Fahriye Tuğba Köş Kübranur Gümüşlü Selin Aktürk Esen Mücella Şahin Mehmet Ali Nahit Şendur Doğan Uncu Predicting hepatocellular carcinoma survival with artificial intelligence Scientific Reports Artificial intelligence Hepatocellular carcinoma Machine learning Survival prediction |
| title | Predicting hepatocellular carcinoma survival with artificial intelligence |
| title_full | Predicting hepatocellular carcinoma survival with artificial intelligence |
| title_fullStr | Predicting hepatocellular carcinoma survival with artificial intelligence |
| title_full_unstemmed | Predicting hepatocellular carcinoma survival with artificial intelligence |
| title_short | Predicting hepatocellular carcinoma survival with artificial intelligence |
| title_sort | predicting hepatocellular carcinoma survival with artificial intelligence |
| topic | Artificial intelligence Hepatocellular carcinoma Machine learning Survival prediction |
| url | https://doi.org/10.1038/s41598-025-90884-6 |
| work_keys_str_mv | AT ismetseven predictinghepatocellularcarcinomasurvivalwithartificialintelligence AT doganbayram predictinghepatocellularcarcinomasurvivalwithartificialintelligence AT hilalarslan predictinghepatocellularcarcinomasurvivalwithartificialintelligence AT fahriyetugbakos predictinghepatocellularcarcinomasurvivalwithartificialintelligence AT kubranurgumuslu predictinghepatocellularcarcinomasurvivalwithartificialintelligence AT selinakturkesen predictinghepatocellularcarcinomasurvivalwithartificialintelligence AT mucellasahin predictinghepatocellularcarcinomasurvivalwithartificialintelligence AT mehmetalinahitsendur predictinghepatocellularcarcinomasurvivalwithartificialintelligence AT doganuncu predictinghepatocellularcarcinomasurvivalwithartificialintelligence |