BackgroundPolycystic ovarian syndrome [PCO] is a worldwide endocrine disorder affecting women of reproductive age. Diagnosis and differentiation of PCO phenotypes are crucial for disease prognosis, fertility outcome, and treatment planning. This study aims to assess arterial spin labeling perfusion (ASL) and diffusion-weighted imaging (DWI) derived metrics in the diagnosis of PCO, differentiation of its phenotypes, and correlation of these metrics with laboratory measurements.ResultsASL and DWI of the pelvis were performed on 72 PCO patients and another 20 age-matched control group. Two observers measured the blood flow (BF) and ADC in the ovarian stroma. Serum levels of testosterone, dehydroepiandrosterone sulfate (DHEAS), and body mass index (BMI) were calculated. BF values were significantly higher in PCO patients than in control cases (P = 0.001), with area under the curve (AUC) of (0.94 and 0.89) and accuracy of (96% and 92%) for both observers, respectively. Also, BF values were significantly higher in classic than in non-classic PCO cases (P = 0.001), with AUC of (0.92 and 0.90) and accuracy of (91%) for both observers, respectively. ADC values were significantly lower in PCO patients than in control cases (P = 0.001), with AUC of (0.85 and 0.84) for the first observer and second observer, respectively. ADC values were significantly lower in classic PCO patients than in non-classic patients (P = 0.001), with AUC of (0.85 and 0.84) and accuracy of (77% and 81%) for both observers, respectively. Combined values of BF and ADC showed an accuracy of 91% and 86% for differentiating patient from control cases for both observers, respectively, and an accuracy of 92% for differentiating classic from non-classic PCO phenotypes. A significant correlation was found between ADC, BF metrics, and both serum testosterone and DHEAS levels (P < 0.05).ConclusionsCombination of ASL and ADC can be used in PCO diagnosis and can help in the differentiation of its phenotypes. Serum levels of testosterone and DHEAS have a significant correlation with ADC and BF metrics.
Read full abstract