Title: Integrating pathological images and genomics data to identify prognostic features related to recurrence of sarcoma

Authors: Zengxin Li; Shiling Song; Jin Deng

Addresses: College of Mathematics and Informatics, South China Agricultural University, Guangzhou 510642, China ' College of Mathematics and Informatics, South China Agricultural University, Guangzhou 510642, China ' College of Mathematics and Informatics, South China Agricultural University, Guangzhou 510642, China

Abstract: To investigate the prognostic prediction of sarcoma recurrence by combining pathological images and genomic data, and to explore potential markers of sarcoma recurrence. Pathological images and genomic data were used for recurrence feature extraction, followed by screening for survival-related features, and finally, pathological images and gene expression data were integrated to analyse prognostic prediction and identify factors affecting patients' survival by using Kaplan-Meier survival curves and Lasso-Cox regression models. Combining pathologic images and genomic data provided better prognostic prediction of patients, and six features were highly associated with sarcoma recurrence. These features have the potential to be key targets for studying sarcoma recurrence and provide valuable insights into personalised treatment of sarcoma recurrence.

Keywords: sarcoma; recurrence; pathological images; bioinformatics; gene expression analysis.

DOI: 10.1504/IJDMB.2025.143042

International Journal of Data Mining and Bioinformatics, 2025 Vol.29 No.1/2, pp.223 - 237

Received: 13 Sep 2023
Accepted: 17 Jul 2024
Published online: 02 Dec 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article