Post written by Jianbo Ni, MD, Yuqing Mao, MD, and Baiwen Li, MD, PhD, from the Department of Gastroenterology, Shanghai General Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China.

Our multicenter study demonstrates that linear-array EUS (L-EUS) outperformed MRCP in diagnosing pancreas divisum (PD), particularly in symptomatic patients. With a sensitivity of 90.8% and an area under the receiver operating characteristic curve of 95.7%, L-EUS proved to be a highly reliable modality that could complement or even precede MRCP in diagnostic algorithms, especially when performed by experienced endosonographers.

PD remains an under-recognized contributor to idiopathic recurrent acute pancreatitis and chronic pancreatitis. Although MRCP is widely used as a noninvasive tool, its diagnostic accuracy is limited in subtle ductal variants and minor papilla anomalies. In many real-world settings, secretin is not routinely available. Our findings highlight that L-EUS, when standardized and performed with adequate expertise, can provide high-resolution, dynamic imaging with superior sensitivity, allowing earlier and more-accurate identification of PD.

Our data suggest that L-EUS could serve as a first-line diagnostic modality in patients with high clinical suspicion of PD, potentially reducing the need for more-invasive ERCP-based diagnostic procedures. We also underscore the importance of standardizing EUS imaging protocols and diagnostic criteria—such as the “stack sign” and ventral-dorsal transition—to ensure reproducibility across institutions. By incorporating L-EUS into routine evaluation, clinicians may achieve more-precise diagnoses, reduce adverse events, and tailor therapeutic strategies more effectively.

We are focused on standardizing interpretation of key EUS features in PD, such as the stack sign and ductal transitions. To improve diagnostic consistency across centers, we are developing concise training modules for endosonographers.

In parallel, we are exploring artificial intelligence—assisted EUS image analysis to support real-time feature recognition and reduce interobserver variability. Together, these efforts aim to enhance the accuracy, reproducibility, and accessibility of L-EUS in PD diagnosis, particularly at centers with limited experience.

Graphical abstract

Read the full article online.

The information presented in Endoscopedia reflects the opinions of the authors and does not represent the position of the American Society for Gastrointestinal Endoscopy (ASGE). ASGE expressly disclaims any warranties or guarantees, expressed or implied, and is not liable for damages of any kind in connection with the material, information, or procedures set forth.