Improving hepatocellular carcinoma diagnosis using an ensemble classification approach based on Harris Hawks Optimization

LiuRen Lin; YunKuan Liu; Min Gao; Amin Rezaeipanah

doi:10.1016/j.heliyon.2023.e23497

Improving hepatocellular carcinoma diagnosis using an ensemble classification approach based on Harris Hawks Optimization

Heliyon. 2023 Dec 9;10(1):e23497. doi: 10.1016/j.heliyon.2023.e23497. eCollection 2024 Jan 15.

Authors

LiuRen Lin¹, YunKuan Liu², Min Gao³, Amin Rezaeipanah⁴

Affiliations

¹ Department of Pharmacy and Machinery, Qujing Second People's Hospital, Yunnan, Qujing, 655000, China.
² Yunnan University of Chinese Medicine, Yunnan Key Laboratory of External Drug Delivery System and Preparation Technology in Universities, Yunnan, Kunming, 650500, China.
³ Faculty of Life Science and Technology, Kunming University of Science and Technology, Yunnan, Kunming, 650500, China.
⁴ Department of Computer Engineering, Persian Gulf University, Bushehr, Iran.

Abstract

Hepato-Cellular Carcinoma (HCC) is the most common type of liver cancer that often occurs in people with chronic liver diseases such as cirrhosis. Although HCC is known as a fatal disease, early detection can lead to successful treatment and improve survival chances. In recent years, the development of computer recognition systems using machine learning approaches has been emphasized by researchers. The effective performance of these approaches for the diagnosis of HCC has been proven in a wide range of applications. With this motivation, this paper proposes a hybrid machine learning approach including effective feature selection and ensemble classification for HCC detection, which is developed based on the Harris Hawks Optimization (HHO) algorithm. The proposed ensemble classifier is based on the bagging technique and is configured based on the decision tree method. Meanwhile, HHO as an emerging meta-heuristic algorithm can select a subset of the most suitable features related to HCC for classification. In addition, the proposed method is equipped with several strategies for handling missing values and data normalization. The simulations are based on the HCC dataset collected by the Coimbra Hospital and University Center (CHUC). The results of the experiments prove the acceptable performance of the proposed method. Specifically, the proposed method with an accuracy of 97.13 % is superior in comparison with the equivalent methods such as LASSO and DTPSO.

Keywords: Ensemble classification; Feature selection; HCC; HHO; Medical diagnosis system.