Endoscopic ultrasonography (EUS) is the most accurate imaging modality for the evaluation
of different types of pancreatic cystic lesions. Our aim was to analyze EUS images
of pancreatic cystic lesions using an image processing software. We specified the
echogenicity of the lesions by measuring the gray value of pixels inside the selected
areas. The images were divided into groups (serous cystic neoplasm /SCN/, intraductal
papillary mucinous neoplasms and mucinous cystic neoplasms /Non-SCN/ and Pseudocyst)
according to the pathology results of the lesions. Overall, 170 images were processed
by the software: 81 in Non-SCN, 30 in SCN and 59 in Pseudocyst group. The mean gray
value of the entire lesion in the Non-SCN group was significantly higher than in the
SCN group (27.8 vs. 18.8; p < 0.0005). The area ratio in the SCN, Non-SCN and Pseudocyst
groups was 57%, 39% and 61%, respectively; significantly lower in the Non-SCN group
than in the SCN or Pseudocyst groups (p < 0.0005 and p < 0.0005, respectively). The
lesion density was also significantly higher in the Non-SCN group compared to the
SCN or Pseudocyst groups (4186.6/mm2 vs. 2833.8/mm2 vs. 2981.6/mm2; p < 0.0005 and
p < 0.0005, respectively). The EUS image analysis process may have the potential to
be a diagnostic tool for the evaluation and differentiation of pancreatic cystic lesions.
