You are here

Article Text

Article menu
Download PDFPDF
PostScript
Letter
Artificial intelligence in EUS for autoimmune pancreatitis: bias and real life
  1. Matteo Tacelli,
  2. Piera Zaccari,
  3. Gabriele Capurso,
  4. Paolo Giorgio Arcidiacono
  1. Pancreato-Biliary Endoscopy and EUS Division, Pancreas Translational and Clinical Research Center, San Raffaele Scientific Institute IRCCS, Milano, Italy
  1. Correspondence to Dr Matteo Tacelli, Pancreato-Biliary Endoscopy and EUS Division, Pancreas Translational and Clinical Research Center, San Raffaele Scientific Institute IRCCS, Milano 20132, Italy; matteo.tacelli{at}gmail.com

https://doi.org/10.1136/gutjnl-2021-324338

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    The diagnosis of autoimmune pancreatitis (AIP) is often challenging and distinguishing it from chronic pancreatitis or pancreatic cancer (PC) could be very difficult. In particular, AIP can present at CT scan, MRI or endoscopic ultrasound (EUS) as a focal mass, sometimes with no other specific peculiarities. Furthermore, current international guidelines1 2 consider EUS only as a way to obtain cyto-histological specimens to exclude malignancy, regardless to its diagnostic capacity based on morphologic aspects.

    For these reasons, we read with great interest the paper by Marya et al 3 about the possible role of a new EUS-convolutional neural network (EUS-CNN) model for the diagnosis of AIP. In this model, videos and pictures of EUS, accurately preselected by the endosonographers, were …

    View Full Text

    Footnotes

    • Contributors All the authors collaborated on the elaboration and revision of the manuscript.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; internally peer reviewed.