cs.AISep 15, 2026

Semi-Supervised Learning-Based Genetic Biomarkers Dataset for Multiple-Stage Hepatocellular Carcinoma Prediction

Authors: Ahmed Ammar Kubba, Manar Abu Talib, Jibran Sualeh Muhammad, Ali Bou Nassif, Abdalla Sayed Mohamed, Darko Castven, Jens U. Marquardt

Organizations: Department of Computer Science, University of Sharjah, Sharjah, UAE · Department of Biomedical Sciences, University of Birmingham, Birmingham, United Kingdom · Department of Computer Engineering, University of Sharjah, Sharjah, UAE · Department of Medicine I, University Medical Center Schleswig-Holstein, Campus Lübeck, Lübeck, Germany

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Liver cancer is a complex disease responsible for a high number of deaths across the globe each year, making automated solutions for liver cancer classification urgent. The most common form of liver cancer is hepatocellular carcinoma (HCC), accounting for over 90% of liver cancer cases. There is a distinct lack of publicly available HCC datasets utilizing genomic data, which is necessary for training artificial intelligence (AI) models for automated HCC classification. This study proposes constructing a multi-stage HCC dataset using XGBoost and Semi-Supervised learning on three separate datasets of genomic biomarkers, utilizing their existing labels in the Semi-Supervised learning process to label the proposed dataset. The proposed dataset consists of 770 patient samples in total, categorized into five classes that represent normal tissue alongside different stages of HCC. Each sample in the dataset consists of 11,150 different gene expression levels. The XGBoost model demonstrated a final classification accuracy of 96.5% during the Semi-Supervised learning process.

Figures & tables

Fig. 1: Visual summary of the paper methodology, outlining each step in the process of constructing the HCC dataset.
ClassDescription
Surrounding Liver (SL)This class represents patient tissue samples taken from the surrounding liver area of the patient, which contains gene expressions that indicate no HCC in the patient.
Early Hepatocellular Carcinoma (EHCC)This class represents samples of patients with a diagnosis of early-stage HCC.
Progressed Hepatocellular Carcinoma (PHCC)This class is assigned to patients with a diagnosis of progressed HCC.
Low-Grade Dysplastic Nodules (LGDN)Dysplastic nodules are associated with a higher risk of developing HCC; low-grade dysplastic nodules represent a much lower risk.
High-Grade Dysplastic Nodules (HGDN)High-grade dysplastic nodules indicate a high risk of developing HCC, as represented by this class.
TABLE I: Description of the dataset classes, consisting of SL, EHCC, PHCC, LGDN, and HGDN.
Fig. 2: Illustration of the Semi-Supervised learning process, using three different sources for constructing the HCC dataset.
Fig. 3: Visual summary of the Semi-Supervised Training process, using the XGBoost model to create the HCC dataset.
ClassPrecisionRecallF1-ScoreSupport
ehcc98.9%96.7%97.8%92
hgdn66.7%72.7%69.6%11
lgdn87.5%100%93.3%21
phcc100%100%100%99
sl100%75%85.7%8
Accuracy96.5%96.5%96.5%96.5%
TABLE II: This table records the XGBoost performance metrics, consisting of accuracy, precision, recall, and F1 score on the HCC dataset.
Fig. 4: Visual comparison of the XGBoost model’s performance across the HCC dataset’s five classes.
Fig. 5: Confusion matrix for the XGBoost model, visualizing the misclassifications of the model across the five classes.
Fig. 6: This graph illustrates the frequency distribution of the different classes in the HCC dataset.
Fig. 7: PCA plot of the HCC dataset, illustrating the distribution of the classes in feature space.
More in this topic

Cancer Genomics

Cancer genomics uses molecular data to understand tumor biology, identify biomarkers, and guide diagnosis, prognosis, and treatment.

See all 34 papers

Explore similar work

CardsList

  1. Potential of Artificial Intelligence Algorithms for Identification of Relevant Diagnostic and Prognostic Biomarkers of Early-Stage Liver Cancer

    Sep 14, 2026Ali Bou Nassif, Darko Castven, Manar Abu Talib +4Cancer GenomicsMedical Diagnosis

    Department of Computer Engineering, College of Computing and Informatics University of Sharjah, Sharjah, United Arab Emirates · Department of Medicine I, University Medical Center Schleswig-Holstein, University of Lubeck, Lübeck, Germany · Department of Computer Science, College of Computing and Informatics University of Sharjah, Sharjah, United Arab Emirates +1

  2. AMPLIFAI: A Multiphase CT Dataset for Benchmarking Clinical Reasoning in LI-RADS Assessment of Liver Lesions

    Aug 14, 2026Pranav Kulkarni, Nikhil Shah, Amritansh Suryavanshi +9Medical Image Benchmarks

    University of Maryland Institute for Health Computing, North Bethesda, MD 20852 · Department of Computer Science, University of Maryland, College Park, MD 20742 · Department of Diagnostic Radiology and Nuclear Medicine, University of Maryland School of Medicine, Baltimore, MD 21201 +2

  3. A Digital Pathology Resource for Liver Cancer Quantification with Datasets, Benchmarks, and Tools

    Apr 22, 2026Ying Xiao, Shimiao Tang, Xitong Ling +11Computational PathologyImage Segmentation

    Department of Pathology, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua Medicine, Tsinghua University, Beijing, China · Shenzhen International Graduate School, Tsinghua University, Shenzhen, China