Endometriomas: Diagnostic performance of US

Maitray D. Patel, Vickie A. Feldstein, Dillon C Chen, Scott D. Lipson, Roy A. Filly

Research output: Contribution to journalArticlepeer-review

206 Scopus citations


PURPOSE: To determine the diagnostic performance of specific ultrasonographic (US) features in discriminating endometriomas from other adnexal masses. MATERIALS AND METHODS: Two sonologists independently reviewed the sonograms of 252 adnexal masses in 226 women and recorded US features by using a standardized checklist. The diagnostic performance of specific US features and overall reviewer impression in discriminating endometriomas from other adnexal masses were evaluated. RESULTS: There were 40 endometriomas. Diffuse low-level internal echoes were present in 38 (95%) endometriomas and 40 (19%) nonendometriomas (positive likelihood ratio, 5). The positive likelihood ratio for the diagnosis of endometrioma increased to 8 if masses with neoplastic features at gray-scale US were excluded, allowing identification of 30 endometriomas (75%). The presence of multilocularity or hyperechoic wall foci further increased the positive likelihood ratio to 48, allowing the identification of 18 endometriomas (45%). CONCLUSION: An adnexal mass with diffuse low-level internal echoes and absence of particular neoplastic features is highly likely to be an endometrioma if multilocular ity or hyperechoic wall foci are present. A patient with a mass with diffuse low-level internal echoes and other US features may benefit from additional imaging.

Original languageEnglish (US)
Pages (from-to)739-745
Number of pages7
Issue number3
StatePublished - Jan 1 1999
Externally publishedYes


  • Dermoid
  • Endometriosis
  • Ovary, neoplasms
  • Pelvic organs, abnormalities
  • Pelvic organs, diseases
  • Pelvic organs, neoplasms
  • Pelvic organs, US

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'Endometriomas: Diagnostic performance of US'. Together they form a unique fingerprint.

Cite this