Yuan C, Hu R, Dang SC. Interpretable extracellular vesicle long RNA framework for noninvasive pancreatic cancer diagnosis: A multi-omics artificial intelligence-driven liquid biopsy paradigm. World J Gastrointest Oncol 2026; 18(9): 117360 [DOI: 10.4251/wjgo.117360]
Corresponding Author of This Article
Sheng-Chun Dang, Professor, Department of General Surgery, The Affiliated Hospital of Jiangsu University, No. 438 Jiefang Road, Zhenjiang 212000, Jiangsu Province, China. dscgu@163.com
Research Domain of This Article
Oncology
Article-Type of This Article
review-article
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
World J Gastrointest Oncol. Sep 15, 2026; 18(9): 117360 Published online Sep 15, 2026. doi: 10.4251/wjgo.117360
Interpretable extracellular vesicle long RNA framework for noninvasive pancreatic cancer diagnosis: A multi-omics artificial intelligence-driven liquid biopsy paradigm
Chang Yuan, Rong Hu, Sheng-Chun Dang
Chang Yuan, Sheng-Chun Dang, Department of General Surgery, The Affiliated Hospital of Jiangsu University, Zhenjiang 212000, Jiangsu Province, China
Rong Hu, Department of Geriatrics, Zhenjiang First People’s Hospital, Zhenjiang 212000, Jiangsu Province, China
Co-first authors: Chang Yuan and Rong Hu.
Author contributions: Yuan C and Hu R contributed equally to this article as co-first authors; Yuan C, Hu R, and Dang SC contributed to manuscript writing and performed bibliographic search; Dang SC revised the final paper; and all authors have read and approved the final version of the manuscript.
AI contribution statement: During the preparation of this manuscript, the authors used artificial intelligence tools for language editing (Gemini 3.1) and figure creation (Picdoc). All artificial intelligence-generated content was subsequently reviewed, edited, and validated by the authors, who take full responsibility for the accuracy and integrity of the published article.
Supported by the Social Development Project of Zhenjiang City, No. SH2024061.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
Corresponding author: Sheng-Chun Dang, Professor, Department of General Surgery, The Affiliated Hospital of Jiangsu University, No. 438 Jiefang Road, Zhenjiang 212000, Jiangsu Province, China. dscgu@163.com
Received: December 8, 2025 Revised: January 26, 2026 Accepted: March 2, 2026 Published online: September 15, 2026 Processing time: 263 Days and 12.4 Hours
Abstract
Diagnosing pancreatic ductal adenocarcinoma at an early stage is highly challenging, which largely drives its persistently high mortality rate. Recently, an interpretable machine learning framework known as ECD-itMLF was introduced to tackle this issue. By analyzing plasma extracellular vesicle (EV) long RNAs, the model achieved an impressive area under the curve of 0.9698 for early detection. In this opinion review, we examine this framework's breakthroughs against the broader backdrop of artificial intelligence-driven multi-omics, while also reviewing the latest progress in EV biology and extraction techniques. Crucially, we argue that such models must be strictly tested against common clinical confounders—especially diabetes and obstructive jaundice—before they can be trusted. We also urge a shift toward more actionable clinical metrics, such as positive predictive values and decision curve analysis. Finally, to address modern oncology’s need for intuitive data interpretation, we outline comprehensive multi-omics strategies and integrated diagnostic blueprints. While the ECD-itMLF paradigm shows tremendous promise, we also look ahead to its potential in monitoring minimal residual disease and advancing single-EV technologies, cautioning that rigorous standardization remains mandatory before everyday clinical use.
Core Tip: We examine the ECD-itMLF interpretable machine learning model, which recently achieved a remarkable area under the curve of 0.9698 for early pancreatic ductal adenocarcinoma detection using sparse extracellular vesicle long RNA data. Despite this excellent performance, moving the tool into the clinic requires testing it against common patient variables like diabetes and jaundice. In this article, we break down the model’s explainable artificial intelligence roots and present visual summaries of current multi-omics research. Looking ahead, we discuss how this extracellular vesicle-based approach could be adapted for tracking minimal residual disease and integrating multiple analytes, providing a realistic path from the lab to everyday precision oncology.
