Scholarlink

<p>This paper reviews the global evolution of synthetic data (SD) generation in the field of genomic cancer medicine, with an analysis of research trends from the past decade. The use of artificial intelligence, particularly machine learning and deep learning techniques has transformed this area, providing solutions to overcome the limited availability of real clinical data. Through a bibliometric analysis of a wide sample of scientific articles from SCOPUS, this study highlights the adoption of SD generation techniques in oncological applications, focusing on major methodologies and challenges. Key application areas, such as multi-omics integration (genomics, transcriptomics, and proteomics) and tumor genomic heterogeneity, emerge as fields of growing interest. Despite noise management and performance optimization challenges, advanced machine learning techniques prove essential for generating high-quality SD that reflects biological complexity. The study also identifies key open challenges, such as simulation accuracy and noise control, offering insights into future applications of SD in personalized medicine and cancer therapy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Synthetic data generation in genomic cancer medicine: a review of global research trends in the last ten years

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月15日

查看全文
  • Valentina De Nicoló,
  • Maria Frasca,
  • Agnese Graziosi,
  • Gianluca Gazzaniga,
  • Davide La Torre,
  • Arianna Pani

摘要

This paper reviews the global evolution of synthetic data (SD) generation in the field of genomic cancer medicine, with an analysis of research trends from the past decade. The use of artificial intelligence, particularly machine learning and deep learning techniques has transformed this area, providing solutions to overcome the limited availability of real clinical data. Through a bibliometric analysis of a wide sample of scientific articles from SCOPUS, this study highlights the adoption of SD generation techniques in oncological applications, focusing on major methodologies and challenges. Key application areas, such as multi-omics integration (genomics, transcriptomics, and proteomics) and tumor genomic heterogeneity, emerge as fields of growing interest. Despite noise management and performance optimization challenges, advanced machine learning techniques prove essential for generating high-quality SD that reflects biological complexity. The study also identifies key open challenges, such as simulation accuracy and noise control, offering insights into future applications of SD in personalized medicine and cancer therapy.