Association of Choroid Plexus Dysfunction and Cognitive Decline in Preeclampsia: Using <scp>T1WI</scp> Imaging, Quantitative Susceptibility Mapping and Deep‐Learning‐Based Segmentation

Preeclampsia is a severe pregnancy complication that can cause brain injury, yet early detection of related cognitive deficits remains challenging. Therefore, in order to investigate alterations in choroid plexus volume (CPV) and susceptibility values of the choroid plexus (ChP) obtained from quantitative susceptibility mapping (QSM) in preeclampsia patients, we enrolled 281 participants, comprising 98 nonpregnant healthy controls (NPHC), 85 pregnant healthy controls (PHC), and 98 patients with preeclampsia. All participants were scanned on a 1.5 T MR scanner. The results of clinical characteristics and cognitive tests were collected from all the participants. One-way ANOVA tests were used to analyze the differences in CPV and susceptibility values of ChP among the three groups. Multiple linear regression analysis was used to find the factors that influenced CPV and its susceptibility values, as well as cognitive decline. Additionally, receiver operating characteristic (ROC) analysis was employed to evaluate the diagnostic performance of the two imaging measures. Preeclampsia patients exhibited smaller CPV and higher susceptibility values compared to the other groups (p < 0.001; p < 0.001). Significant negative correlations were observed between body mass index (BMI), mean arterial pressure and CPV/eTIV (β = -0.100, 95% CI = -0.158 ~ -0.042, p = 0.001; β = -0.022, 95% CI = -0.033 ~ -0.011, p < 0.001). Additionally, significant positive correlations were observed between BMI (β = 0.455, 95% CI = 0.125 ~ 0.786, p = 0.007), mean arterial pressure (β = 0.170, 95% CI = 0.107 ~ 0.232, p < 0.001), hemoglobin (β = 0.152, 95% CI = 0.051 ~ 0.254, p = 0.003) and susceptibility values of ChP. Furthermore, CPV/eTIV and susceptibility values of ChP could be independent contributing factors of scores of TMT. The combination of CPV, susceptibility values of ChP, BMI and gestational week could distinguish preeclampsia from pregnant groups (AUC = 0.787, 95% CI = 0.722-0.853, p < 0.001) as well as distinguish individuals with cognitive decline from preeclampsia patients (AUC = 0.737, 95% CI = 0.621-0.844, p < 0.001). These findings indicate that smaller CPV and higher susceptibility values characterize preeclampsia and may serve as auxiliary indices for its diagnosis and related cognitive decline.