Back to the list
Congress: ECR25
Poster Number: C-28431
Type: Poster: EPOS Radiologist (scientific)
DOI: 10.26044/ecr2025/C-28431
Authorblock: I. Gruzdev, K. Zamyatina, A. Ustalov, S. A. Shmeleva, V. Aznaurov, V. Gurina, A. Mazurok, B. M. Karteev, E. V. Kondratyev; Moscow/RU
Disclosures:
Ivan Gruzdev: Nothing to disclose
Kseniia Zamyatina: Nothing to disclose
Andrey Ustalov: Nothing to disclose
Sofiia Antonovna Shmeleva: Nothing to disclose
Vladimir Aznaurov: Nothing to disclose
Vera Gurina: Nothing to disclose
Alina Mazurok: Nothing to disclose
Batnasan Ming'Yanovich Karteev: Nothing to disclose
Evgeny V Kondratyev: Nothing to disclose
Keywords: Abdomen, CT, Computer Applications-3D, Computer Applications-General, Segmentation, Cancer, Tissue characterisation
Purpose Segmentation is an important process for AI radiomics, radiotherapy planning, assessing the dynamics of the pathological process, and radiomics. Texture analysis is a technique with high potential for use in clinical diagnostic practice. Texture analysis uses images obtained during routine diagnostic procedures, but involves a group of mathematical calculations performed on the information contained within the images [1]. This method is especially useful in the analysis of CT studies of the pancreas, since classical medical imaging methods face some significant...
Read more Methods and materials To determine the permissibility of errors in segmentation, 3 radiologists with experience in abdominal imaging and marking pancreatic pathology made segmentations of 30 contrast-enhanced CT studies of the pancreas, performed both at the A.V. Vishnevsky National Medical Research Center for Surgery, and various other medical institutions on CT-scanners from various manufacturers. Of the 30 studies performed, 15 patients had a normal pancreas, the remaining 15 patients had various pancreatic lesions (ductal and periductal adenocarcinoma, neuroendocrine tumor, chronic pancreatitis). [fig 1] [fig 2] To achieve...
Read more Results In all cases of segmentation the following was observed: the threshold of 0.9 (excellent reliability) for the arterial phase of the study was not achieved when the border of the pancreas was hyposegmented by more than 2 mm and hypersegmented by more than 1 mm; for the portal phase of the study the threshold is 0.9 was not achieved when the border of the pancreas was hyposegmented by more than 3 mm and hypersegmented by more than 2 mm. Significant...
Read more Conclusion Minor errors in pancreas segmentation do not affect the reproducibility of the texture features, which allows to obtain accurate results using texture analysis. Arterial phase study segmentation is more susceptible to errors due to its more heterogeneous structure, so more accurate segmentation is required. But if the radiologist has doubts about determining the boundaries of the pancreatic parenchyma, hyposegmentation of the organ is preferable. Hypersegmentation reduces the reproducibility of the resulting texture features.
Read more References 1. Castellano G, Bonilha L, Li LM, Cendes F. Texture analysis of medical images. Clin Radiol. 2004 Dec;59(12):1061-9. doi: 10.1016/j.crad.2004.07.008. PMID: 15556588.2. Awe AM, Rendell VR, Lubner MG, Winslow ER. Texture Analysis: An Emerging Clinical Tool for Pancreatic Lesions. Pancreas. 2020 Mar;49(3):301-312. doi: 10.1097/MPA.0000000000001495. PMID: 32168248; PMCID: PMC7135958.3. Renard F, Guedria S, Palma N, Vuillerme N. Variability and reproducibility in deep learning for medical image segmentation. Sci Rep. 2020 Aug 13;10(1):13724. doi: 10.1038/s41598-020-69920-0. PMID: 32792540; PMCID: PMC7426407.4. Adelsmayr G, Janisch...
Read more
GALLERY