H Schrader 1 , M Wiese , M Ellrichmann , O Belyaev , W Uhl , A Tannapfel , W Schmidt , J Meier . Show Affiliations »
Abstract
PURPOSE: EUS elastography has been used to facilitate the diagnosis of pancreatic cancer, but as yet the interpretation of this procedure has been largely subjective. The present study has been designed to validate a quantitative approach for the analysis of EUS elastography, and to assess its relationship with pancreatic fibrosis. MATERIALS AND METHODS: 86 patients with malignant pancreatic masses and 28 control subjects without any evidence of pancreatic diseases were examined by EUS elastography. EUS video sequences were subjected to a quantitative analysis based on mean hue histogram analysis. Pancreatic fibrosis was determined by quantitative morphometry in tissue specimens from 36 patients. RESULTS: The mean RGB (red, green, blue) value was significantly higher in the cancer patients compared to the controls (14.0 ± 0.4 vs. 11.5 ± 0.9; p = 0.0085), albeit with significant overlap between the groups. In contrast, a much sharper separation between the groups was obtained based on the individual color values for blue, green and red (p < 0.0001, respectively). By these means, 100 % sensitivity and specificity for the distinction between tumor and normal tissue was obtained for the blue color value, while the red and green color values were less discriminative. The fractional fiber content of the tumors was unrelated to the respective hue histogram color values. CONCLUSION: Quantitative EUS elastography allows for clear differentiation between malignant pancreatic tumors and normal tissue. Using this approach, we demonstrated that the stiffness of pancreatic tumors is largely independent of their fiber content. © Georg Thieme Verlag KG Stuttgart · New York.
PURPOSE: EUS elastography has been used to facilitate the diagnosis of pancreatic cancer , but as yet the interpretation of this procedure has been largely subjective. The present study has been designed to validate a quantitative approach for the analysis of EUS elastography, and to assess its relationship with pancreatic fibrosis . MATERIALS AND METHODS: 86 patients with malignant pancreatic masses and 28 control subjects without any evidence of pancreatic diseases were examined by EUS elastography. EUS video sequences were subjected to a quantitative analysis based on mean hue histogram analysis. Pancreatic fibrosis was determined by quantitative morphometry in tissue specimens from 36 patients . RESULTS: The mean RGB (red, green, blue) value was significantly higher in the cancer patients compared to the controls (14.0 ± 0.4 vs. 11.5 ± 0.9; p = 0.0085), albeit with significant overlap between the groups. In contrast, a much sharper separation between the groups was obtained based on the individual color values for blue, green and red (p < 0.0001, respectively). By these means, 100 % sensitivity and specificity for the distinction between tumor and normal tissue was obtained for the blue color value, while the red and green color values were less discriminative. The fractional fiber content of the tumors was unrelated to the respective hue histogram color values. CONCLUSION: Quantitative EUS elastography allows for clear differentiation between malignant pancreatic tumors and normal tissue. Using this approach, we demonstrated that the stiffness of pancreatic tumors is largely independent of their fiber content. © Georg Thieme Verlag KG Stuttgart · New York.
Entities: Disease
Species
Mesh: See more »
Year: 2011
PMID: 21630184 DOI: 10.1055/s-0031-1273256
Source DB: PubMed Journal: Ultraschall Med ISSN: 0172-4614 Impact factor: 6.548