Placenta plays a crucial role in ensuring proper fetal growth during pregnancy. However, placental insufficiency (PI), where the placenta fails to provide enough oxygen and nutrients to the fetus, can lead to adverse pregnancy outcomes. Magnetic Resonance Imaging (MRI) is a non-invasive imaging modality that can quantify placental blood flow and detect PI. In this study, we aimed to develop a new method for in vivo quantification of placental insufficiency using BOLD MRI and evaluate its diagnostic accuracy.
Methodology
We recruited pregnant women at 18-24 weeks gestation and acquired BOLD MRI scans at multiple time points during pregnancy. We developed a deep learning model to quantify placental blood flow and detected PI based on the ratio of the placental to cerebral blood volume. We evaluated the diagnostic accuracy of our method using a combination of MRI and clinical data.
Results
Our results showed that our BOLD MRI-based method accurately detects PI in vivo with high sensitivity and specificity. Compared to conventional methods, our approach provided more accurate quantification of placental blood flow and reduced the need for invasive measurements. We observed a strong correlation between our MRI-based method and clinical data, suggesting that MRI can be used as a non-invasive biomarker for PI diagnosis.
Discussion
Our study demonstrates the feasibility of in vivo quantification of placental insufficiency using BOLD MRI. Our approach provides accurate and reliable measurements of placental blood flow, which can be used to diagnose PI non-invasively. The use of deep learning algorithms enables us to accurately detect PI in vivo, even at early stages of pregnancy. These results have important implications for improving prenatal care and reducing the risk of adverse pregnancy outcomes.
Conclusion
In conclusion, our study demonstrates the potential of BOLD MRI-based deep learning methods for in vivo quantification of placental insufficiency. Our approach provides accurate and reliable measurements of placental blood flow, which can be used to diagnose PI non-invasively. These results have important implications for improving prenatal care and reducing the risk of adverse pregnancy outcomes. Future studies will focus on further validating our method in larger populations and exploring its clinical applications.
