Original scientific paper
Prediction of Ovarian Tumor Malignancy
Hrvojka Soljačić Vraneš
Petar Klarić
Zdenko Sonicki
Vesna Gall
Marija Jukić
Ante Vuković
Abstract
Ovarian cancer is the leading cause of mortality among gynecological cancers. The aim of the study was to form the
decision rules for distinguishing benign from malignant ovary lesions. The research was conducted on 201 women with
ovary tumor. Commonly used specific markers for ovarian cancer (biochemical marker Ca 125, ultrasound and vascular
markers) were used. The signifficant differenece in the presence of an ultrasound and vascular markers between benign
and malignant ovary changes along with the signifficantly different level of Ca 125 is confirmed. To a specific marker
certain score number was appointed and the scoring system was formed. The incidence of benign/malignant ovary changes
was observed in the researched group regarding anthropometric parameters (age, marital and menopausal status
and number of deliveries). There is also signifficant difference in the incidence of benign/malignant ovary tumor regarding
these parameters. Based on combination of the scoring system and anthropometric parameters the decision rules for
distinguishing benign from malignant ovary tumors were formed. The logistic regression method was used. We proved
that this method has higher accuracy in prediction of malignancy in women with ovary tumors than using morphological,
doppler or anthropometric parameters separately.
Keywords
Hrčak ID:
72171
URI
Publication date:
30.9.2011.
Visits: 1.674 *