AI-powered liquid biopsy for early detection of gastrointestinal cancers

Abstract

Gastrointestinal cancers (GICs) are a leading cause of cancer-related mortality worldwide, largely due to late-stage diagnosis. Liquid biopsy has emerged as a promising non-invasive diagnostic tool, utilizing circulating tumor DNA (ctDNA), circulating tumor cells, exosomal RNA (exoRNA), and tumor-educated platelets for early cancer detection. Challenges, including data complexity, low biomarker abundance, and detection variability, require advanced computational solutions. Artificial intelligence (AI), particularly machine learning (ML) and deep learning (DL), has significantly improved the accuracy and clinical utility of liquid biopsy by enabling high-throughput biomarker discovery, multi-omics integration, and predictive modelling. AI-driven algorithms have enhanced ctDNA mutation profiling, methylation analysis, and fragmentomics, offering superior sensitivity and specificity for early GIC detection. Additionally, AI-based analysis of exoRNA and platelet-derived biomarkers provides novel insights into tumor progression and patient stratification. Despite these advances, key challenges remain, including data standardization, bias mitigation, and regulatory validation. The implementation of federated learning and ethical AI frameworks can further refine AI-powered liquid biopsy models, paving the way for precision oncology applications. This review highlights advances in AI-powered liquid biopsy for early GIC detection, emphasizing its potential and the need for validation, collaboration, and regulatory alignment for clinical adoption.