Phase-inversion tissue harmonic imaging compared with conventional B-mode ultrasound in the evaluation of pancreatic lesions

• Published in: European radiology Vol. 14; no. 6; pp. 1109 - 1117
• Main Authors: Hohl, Christian, Schmidt, Thorsten, Haage, Patrick, Honnef, Dagmar, Blaum, Marcus, Staatz, Gundula, Guenther, Rolf W
• Format: Journal Article
• Language: English
• Published: Germany Springer Nature B.V 01-06-2004
• Subjects: Adult; Aged; Aged, 80 and over; Female; Humans; Magnetic Resonance Imaging; Male; Middle Aged; Pancreas; Pancreatic Diseases - diagnosis; Pancreatic Diseases - diagnostic imaging; Pancreatic Neoplasms - diagnostic imaging; Pancreatitis - diagnostic imaging; Prospective Studies; Sensitivity and Specificity; Tomography, X-Ray Computed; Ultrasonography - methods
• ISSN: 0938-7994; 1432-1084
• DOI: 10.1007/s00330-003-2191-2

The aim of this study was to compare the diagnostic sensitivity, specificity, and image quality of conventional B-mode US (BM) and phase-inversion tissue harmonic imaging (PTHI) regarding pancreatic pathology. In a prospective study, 107 patients, aged between 28 and 85 years, underwent US examinations of the pancreas with both BM and PTHI in a randomly chosen order. As diagnostic reference, either contrast-enhanced CT or MRI examinations of the upper abdomen were obtained in all patients. Sensitivity and specificity were evaluated using the Student's t test. Differences in overall image quality, lesion conspicuity, fluid-solid differentiation, and delineation of the pancreatic tail were analyzed using Wilcoxon's signed ranks test and Bowker's symmetry test. Sixteen of 107 examined patients (15%) were non-diagnostic and excluded due to technical limitations such as abdominal gas. A total of 60 pancreatic lesions (cysts, acute pancreatitis, dilatation of the pancreatic duct, calcifications, and solid tumors) were diagnosed by CT or MRI. Phase-inversion tissue harmonic imaging had a higher sensitivity of 70% (14 of 20) than BM (60%; 24 of 40) for the detection of pancreatic lesions; however, the difference was not statistically significant ( p=0.46). In the assessment of lesions <1 cm of size, PTHI had a sensitivity of 70% and BM 46.7%, whereby the difference again was not statistically significant. Phase-inversion tissue harmonic imaging proved to be superior to BM regarding overall image quality ( p<0.0001), lesion conspicuity ( p=0.0045), and fluid-solid differentiation ( p=0.0002), as well as the delineation of the pancreatic tail ( p<0.0001). These differences were statistically significant. The statistically significant improvement of image quality with regards to lesion conspicuity, fluid-solid differentiation, and delineation of the pancreatic tail favors the use of PTHI when evaluating the pancreas with US. Sensitivity for pancreatic lesions is increased with PTHI in comparison with conventional sonography (BM), especially in lesions <1 cm in diameter, although the difference was not statistically significant.