Chronic Pancreatitis
Chronic Pancreatitis
Differentiation between pancreatic cancer and chronic pancreatitis with imaging tests is improving but remains imperfect. To improve interpretation of endoscopic ultrasonography (EUS) images, we first reported and demonstrated that an artificial neural network (ANN) analysis of EUS images differentiated between pancreatic malignancy and chronic pancreatitis with similar accuracy compared with blinded videotape analysis and real time analysis (80, 83 and 85%, respectively). Since then advances in EUS technology such as EUS elastography have occurred and this past year Saftoiu et al. evaluated an ANN of EUS elastography in the diagnosis of focal pancreatic masses in a prospective, blinded study of 258 patients (211 with pancreatic cancer and 47 with chronic pancreatitis) from 13 centers in Europe. They reported an overall accuracy of 84% and a significantly higher area under the receiver operating characteristic (ROC) curve compared with mean hue histogram analysis (0.94 vs. 0.85). Thus, this study provides more evidence that ANN analysis of EUS images may improve diagnosis.
Differentiation Between Pancreatic Cancer and Chronic Pancreatitis
Differentiation between pancreatic cancer and chronic pancreatitis with imaging tests is improving but remains imperfect. To improve interpretation of endoscopic ultrasonography (EUS) images, we first reported and demonstrated that an artificial neural network (ANN) analysis of EUS images differentiated between pancreatic malignancy and chronic pancreatitis with similar accuracy compared with blinded videotape analysis and real time analysis (80, 83 and 85%, respectively). Since then advances in EUS technology such as EUS elastography have occurred and this past year Saftoiu et al. evaluated an ANN of EUS elastography in the diagnosis of focal pancreatic masses in a prospective, blinded study of 258 patients (211 with pancreatic cancer and 47 with chronic pancreatitis) from 13 centers in Europe. They reported an overall accuracy of 84% and a significantly higher area under the receiver operating characteristic (ROC) curve compared with mean hue histogram analysis (0.94 vs. 0.85). Thus, this study provides more evidence that ANN analysis of EUS images may improve diagnosis.
Source...