Enhancing Clear-Cell Renal Cell Carcinoma Survival Prediction Using Explainable Machine Learning

Fawad Ahmad, Abu Dhabi University
Heba Ismail, Zayed University

Document Type

Conference Proceeding

Source of Publication

Lecture Notes in Networks and Systems

Publication Date

11-20-2025

Abstract

Clear-cell renal cell carcinoma (ccRCC) is a prevalent subtype of renal cell carcinoma, a life-threatening kidney cancer. The ability to predict patient survival in ccRCC is of utmost importance for personalized treatment strategies and improving patient outcomes. In this thesis, we aim to develop a gene expression-based model for survival prediction in ccRCC, leveraging publicly available datasets such as The Cancer Genome Atlas (TCGA) and the E-MTAB-1980 dataset. By analyzing gene expression data from these datasets, we will utilize machine learning techniques to build a predictive model. However, existing methods face challenges in achieving accurate classification performance, and they work in a black-box manner, thus unable to explain their prediction. To overcome these limitations, we propose explainable machine learning algorithms to provide insights into the model’s decision-making process, enhancing its clinical applicability. Experimental results show that the Random Forest (RF) classifier achieved the highest performance on the TCGA dataset with an accuracy of 0.776 and an F1-score of 0.762. While on the E-MTAB dataset, Multi-layer Perceptron (MLP) outperformed others with an accuracy of 0.802, an F1-score of 0.753, and the highest AUC of 0.980. SHAP analysis revealed important features like age and ZIC2 for TCGA and ANAPC5 for E-MTAB, influencing patient survival outcomes. Our research aims to advance the field of cancer research by improving the understanding and application of ML techniques for survival prediction in ccRCC.

ISBN

[9789819692415]

ISSN

2367-3370

Publisher

Springer Nature Singapore

Volume

1537 LNNS

First Page

629

Last Page

642

Disciplines

Computer Sciences | Medicine and Health Sciences

Keywords

Clear-cell renal cell carcinoma, Clinical applicability, Machine learning

Scopus ID

105023287797

Indexed in Scopus

yes

Open Access

no

Link to Full Text

Share

COinS