Foucauld Chamming's1, Chloé Hangard2, Jean-Luc Gennisson3, Caroline Reinhold4, Laure S Fournier5. 1. Department of Radiology, Institut Bergonié, 229 cours de l'Argonne, Bordeaux 33076, France; Department of Radiology, Université Paris Descartes Sorbonne Paris Cité, Assistance Publique-Hôpitaux de Paris, Hôpital Européen Georges Pompidou, Paris, France. Electronic address: f.chammings@bordeaux.unicancer.fr. 2. Department of Radiology, Université Paris Descartes Sorbonne Paris Cité, Assistance Publique-Hôpitaux de Paris, Hôpital Européen Georges Pompidou, Paris, France. 3. Institut Langevin-Ondes et Images, ESPCI ParisTech CNRS UMR 7587 INSERM U979, Paris, France. 4. Department of Radiology, McGill University Health Center, Montréal, QC, Canada. 5. Department of Radiology, Université Paris Descartes Sorbonne Paris Cité, Assistance Publique-Hôpitaux de Paris, Hôpital Européen Georges Pompidou, Paris, France; Cardiovascular Research Center - PARCC, Université Paris Descartes Sorbonne Paris Cité, UMR-S970, Paris, France.
Abstract
PURPOSE: To investigate the diagnostic accuracy of applying four levels of manual pressure in Shear Wave Elastography (SWE) of the breast and to assess inter-rater reliability. MATERIALS AND METHODS: Single-center prospective preliminary study including patients receiving ultrasound examination of breast lesions as part of routine clinical practice. SWE was performed on 60 breast masses (26 benign and 34 malignant) in 54 patients by a breast fellowship trained radiologist. Stiffness values were compared between benign and malignant masses at four levels of manual compression: none, mild, moderate, and marked. Accuracy of SWE was assessed using receiving operating characteristics analysis at each level. In 18 patients, a second radiologist repeated the SWE acquisitions to evaluate reproducibility. Reproducibility was assessed using intraclass correlation coefficient. RESULTS: Without compression, we observed no significant difference in stiffness (p > 0.99) between benign and malignant lesions, and SWE demonstrated low accuracy (area under the curve = 0.64). Stiffness was higher in malignant lesions at all levels of compression (p < 0.001). SWE demonstrated good accuracy at all three levels of compression (from area under the curve = 0.71 to 0.84 across Emax and Emean), with high interobserver agreement. CONCLUSION: This preliminary study suggests that not using compression during SWE for breast lesion characterization offers suboptimal results. On the contrary, application of compression yields high diagnostic performance with good interobserver agreement and, as such, should be included in routine clinical practice.
PURPOSE: To investigate the diagnostic accuracy of applying four levels of manual pressure in Shear Wave Elastography (SWE) of the breast and to assess inter-rater reliability. MATERIALS AND METHODS: Single-center prospective preliminary study including patients receiving ultrasound examination of breast lesions as part of routine clinical practice. SWE was performed on 60 breast masses (26 benign and 34 malignant) in 54 patients by a breast fellowship trained radiologist. Stiffness values were compared between benign and malignant masses at four levels of manual compression: none, mild, moderate, and marked. Accuracy of SWE was assessed using receiving operating characteristics analysis at each level. In 18 patients, a second radiologist repeated the SWE acquisitions to evaluate reproducibility. Reproducibility was assessed using intraclass correlation coefficient. RESULTS: Without compression, we observed no significant difference in stiffness (p > 0.99) between benign and malignant lesions, and SWE demonstrated low accuracy (area under the curve = 0.64). Stiffness was higher in malignant lesions at all levels of compression (p < 0.001). SWE demonstrated good accuracy at all three levels of compression (from area under the curve = 0.71 to 0.84 across Emax and Emean), with high interobserver agreement. CONCLUSION: This preliminary study suggests that not using compression during SWE for breast lesion characterization offers suboptimal results. On the contrary, application of compression yields high diagnostic performance with good interobserver agreement and, as such, should be included in routine clinical practice.
Authors: Jin You Kim; Jin Joo Kim; Lee Hwangbo; Hie Bum Suh; Ji Won Lee; Nam Kyung Lee; Ki Seok Choo; Suk Kim Journal: Br J Radiol Date: 2021-10-05 Impact factor: 3.039