A risk of malignancy index incorporating CA 125, ultrasound and menopausal status for the accurate preoperative diagnosis of ovarian cancer

The preoperative diagnosis of borderline ovarian tumors: a review of current literature

“…The scoring system that did best when applied on the IOTA data was the risk malignant index of Jacobs et al [75] that correlate menopausal status, serum marker and ultrasonographic findings.…”

Using Bayesian neural networks with ARD input selection to detect malignant ovarian masses prior to surgery

“…Some clinicians prefer not to measure CA125 levels if not strictly necessary. Next, two subjective measurements were selected by some models: the colour score of intratumoural blood flow and the presence of pelvic pain; the former because the expert had to rate the level of intratumoural blood flow on a four-point scale (1)(2)(3)(4), the latter because the patient had to decide whether she felt pelvic pain or not. For benign cases, the average colour score ranged from 1.3 for Milan (Italy) to 2.8 for Malmö (Sweden).…”

“…Note, however, that we used cutoffs based on the final model's performance on the original training set (see first part of Table 3), so that they cannot be used blindly in a resampling framework, which uses several different training sets. A widely used model in practice is the risk of malignancy index (RMI) [3]. Since CA125 levels were not always available, the AUROC for the RMI (0.870, SE = 0.028; see Table 3) could be computed on the test set cases with CA125 information only (n = 236, 76%).…”

“…The results were compared to the widely used RMI scoring index [3], the LR model developed on the same data [21], and a linear benchmark model consisting of a perceptron fitted using the evidence framework. Another important aspect of the evaluation of probabilistic classifiers is to check whether the probabilities are calibrated: if the model outputs a probability of malignancy of 30%, is that probability correct?…”

A new computer-aided diagnostic tool for non-invasive characterisation of malignant ovarian masses: results of a multicentre validation study

“…MRI appeared to have similar sensitivity and slightly better specificity than Doppler TVS [6,7,9] but is more expensive and less available while CA125, CT and PET have poorer diagnostic performance than MRI or Doppler TVS [4]. Different risk-of-malignancy indexes, combining menopausal status, CA125 and TVS, were also proposed but, in validation studies, they did not improve performance above that of Doppler TVS alone [10][11][12][13][14].…”