Sonographic predictors of ovarian malignancy

Area of Science:

• Gynecologic Oncology
• Medical Imaging
• Ultrasound Diagnostics

Background:

• Ovarian cancer diagnosis relies on accurate interpretation of imaging findings.
• Transvaginal sonography is a key tool for evaluating pelvic masses.
• Identifying specific sonographic features can improve malignancy prediction.

Purpose of the Study:

• To determine a combination of sonographic features for predicting ovarian malignancy.
• To develop a risk stratification model for ovarian masses using ultrasound.
• To differentiate between benign, borderline, and invasive ovarian cancers.

Main Methods:

• Retrospective analysis of transvaginal sonograms from 249 women with pelvic masses.
• Twelve sonographic features were assessed by a blinded sonologist.
• Stepwise logistic regression identified features associated with malignancy; a risk model was created and validated on a separate testing set.

Main Results:

• Internal vascularity was a high-risk feature, present in 75% of invasive cancers.
• Thick walls or septa without internal vascularity identified additional invasive and borderline tumors.
• Low-risk features were associated with a 4% incidence of malignancy.

Conclusions:

• Specific sonographic features, including internal vascularity and wall/septal thickness, are crucial for predicting ovarian malignancy.
• A three-level risk stratification model can effectively distinguish between benign and malignant ovarian lesions.
• The absence of high or intermediate risk sonographic features indicates a low likelihood of malignancy.