Shuping Weng^{1} and Ruqi Fang^{1}

Aims: To evaluate the effectiveness of diffusion weighted imaging and dynamic contrast enhanced magnetic resonance imaging in qualitative diagnosis of complex ovarian tumors．

Methods: Semi-quantitative parameters of maximal slope of time-signal intensity curve and apparent diffusion coefficient measurements of diffusion-weighted imaging obtained from the tumor solid regions in 65 cases confirmed complex ovarian tumors were done．

Results: A cut-off value of 1．61 × 10－3mm2 /s for ADC was used in differentiating invasive from borderline tumors．Optimal threshold value for maximal slope was 4．58% per second identifying benign and malignant tumors.

Conclusion: Maximal slope value of time-signal intensity curve is the best index in distinguishing malignant from benign ovarian complex tumors，and ADC value of the solid component is the best index in differentiating invasive from borderline ovarian complex tumors．

Introduction: Clinical treatment strategy of ovarian tumors depends on the nature and clinical stage of the tumor. Magnetic resonance imaging plays an important role with its non-invasive and good soft tissue contrast in the diagnosis and staging of ovarian tumors. In recent years, with the continuous innovation of MR technology, MR functional imaging including diffusion weighted imaging (DWI) and dynamic contrast enhanced MRI (DCE-MRI) has been used to the differential diagnosis of benign and malignant tumors in various systems[1]. It can reflect functional information such as the pathophysiology of tumors at the microvascular, cellular, and molecular levels. The aim of this study was to evaluate the effectiveness of diffusion weighted imaging ( DWI)，apparent diffusion coefficient (ADC) value, and time-intensity curve (TIC) type analysis derived from dynamic contrast enhanced MRI (DCE-MRI) in qualitative diagnosis of complex ovarian tumors．

Methods: 65 cases confirmed pathologically complex ovarian tumors were retrospectively analyzed．Magnetic resonance imaging(MRI) examinations included plain scanning，DCE-MRI and DWI. The semi-quantitative parameters of maximal slope of time-signal intensity curve ( TIC) and the apparent diffusion coefficient(ADC) measurements of diffusion-weighted imaging obtained from the tumor solid regions were determined．The results were compared with the histopathological findings，and using receiver operating characteristic analysis method the optimal threshold values for the diagnosis of benign or malignant tumors and the borderline or invasive tumors were determined. Meanwhile, the corresponding sensitivity，specificity，positive predictive values and negative predictive values were calculated．

Results: In this series，the patient’s age ranged from 14 to 58 years，with a mean of 43±2.6 years．Of the 65 patients，benign tumor was seen in 22 and malignant tumor in 43，which included borderline tumor ( n = 21) and invasive tumor ( n = 22)．Statistical analysis showed that the optimal threshold value for maximal slope was 4．58% per second，at this value the sensitivity，specificity，positive predictive values，and negative predictive values were 85．4%，92．3%，95．8% and 69．2%，respectively．Using the same approach，when a cut-off value of 1．61 × 10－3mm2 /s was used for ADC，the sensitivity，specificity，positive predictive values，and negative predictive values in differentiating invasive from borderline tumors were 86．7%，100%，89．5% and 100% respectively．

Conclusion: The maximal slope value of time-signal intensity curve is the best index in distinguishing malignant from benign ovarian complex tumors，and the ADC value of the solid component is the best index in differentiating invasive from borderline ovarian complex tumors．Combination use of conventional MR imaging，DWI and D CE-MRI can improve the diagnosis of ovarian complex tumors．

【Key words】Ovarian tumor; Dynamic contrast-enhanced; Diffusion-weighted imaging; Apparent diffusion coefficient; Magnetic resonance imaging

Figure 1 Left ovarian benign tumor. T2WI fat suppression image (A) showed
a solid mass in the left attachment area, showing a slightly higher signal. ADC
map (B) showed a significantly lower signal with ADC value of 1. 19 × 10 - 3mm2
/ s in the solid area. MS value was 1.58% / s measured by TIC curve (C) line S
curve fitting. Pathology (D) showed ovarian membrane-fibroma (HE × 40).

Figure 2 Right ovarian invasive tumor. T2WI fat suppression image (E) showed
a solid mass in the right accessory area, with a slightly higher signal in the
solid part. A significantly lower signal in the ADC (F) measured ADC value of
1. 28 × 10 - 3 mm2 / s in the solid area. TIC curve (G) line S curve fitting
measured MS value of 52. 9% / s. Pathology (H) showed ovarian serous
cystadenocarcinoma ( HE × 40).

Figure 3 Left ovarian borderline tumor. T2WI fat suppression image (I)
showed a solid tumor in the left accessory area with slightly higher signal in the
solid part. The ADC map (J) is low signal with ADC value measured 2. 19 × 10 -
3mm2 / s. TIC curve (K) line S curve fitting measured MS value of 8. 50% / s. Pathology
(L) is borderline serous papillary cystadenoma (HE × 40)