Keywords: Uterus, Reproductive, Female Reproductive System, Machine Learning, Artificial Intelligence

Motivation: To quantify female reproductive anatomy in MR imaging.

Goal(s): To develop an AI-based solution to segment the regions of interest (RoIs) for the uterine zone, ovaries, pelvic fluid, and detect benign uterine conditions.

Approach: A deep learning based method is applied on a large representative population of 9334 sagittal T2-weighted female pelvis scans to extract normative menstrual cycle- and aging-curves for various RoIs.

Results: Our proposed normative curves define the standard menstrual cycle and aging trends. RoI segmentation, fibroid, and cyst detection models achieve average foreground dice, specificity and accuracy scores of 83.9%, 95.2% and 94.37%, respectively.

Impact: Proposing a robust, precise AI solution for analyzing female reproductive organs on MR imaging, including uterine zones, ovaries, pelvic fluid, and fibroids/cysts. Using this, we define standard aging and menstrual cycle curves for women.

Keywords: Uterus, Uterus, Cesarean scar pregnancy;

Motivation: Dilation and Curettage (D&C) is the preferred treatment for Cesarean scar pregnancy (CSP) patients. However, current methods for preoperative assessment of intraoperative bleeding are limited.

Goal(s): To develop a nomogram using radiomics and MRI-based clinical information to predict massive hemorrhage during D&C in CSP patients.

Approach: 116 CSP patients were enrolled. Radiomics features and clinical variables were analyzed to conduct the radiomics and the clinical models. The nomogram was established by combining Radscore and the selected clinical variables.

Results: Among the three models, the nomogram achieved the highest prediction performance with an AUC of 0.926 for test data.

Impact: This study first developed the MR-based radiomics nomogram to accurately predict massive hemorrhage preoperatively in CSP patients.

Keywords: Placenta, Placenta

Motivation: This study is motivated by the clinical need to understand and monitor early gestational placenta development. 

Goal(s): The goal is to show that multi-parametric placenta imaging during early gestation is a clinical tool to flag possible pregnancies in danger of multiple adverse effects. 

Approach: This study includes data from a previous perspective study where MRI was taken at two separate timepoints during early gestation. Information on volume, perfusion, and oxygenation were evaluated for longitudinal trends as well as changes driven by obesity and fetal sex outcomes.

Results: We show statistically significant differences of longitudinal trends in most of the parameters between study cohorts. 

Impact: The results in this study could allow for clinicians to gain insight on pregnancy development within the early second trimester currently unavailable with the current imaging with Doppler ultrasound. This will allow for overall life improvement for mother and fetus. 

Keywords: Placenta, fMRI, Gestational hypertension, perfusion, placenta, virtual magnetic resonance elastography

Motivation: Detecting placental dysfunction using MRI before placental macrovascular lesions by ultrasound may help the early identification of placental pathologic changes in gestational hypertension (GH).

Goal(s): To explore changes in placental elasticity and perfusion in GH and control groups using MRI.

Approach: Placental elasticity and perfusion were assessed using virtual magnetic resonance elastography (vMRE) and intravoxel incoherent motion (IVIM). The vMRE and IVIM parameters, MR morphologic parameters, and ultrasound and lab test results were compared between the two groups. 

Results: Placental stiffness and perfusion fraction helped distinguish between the two groups, with no substantial differences in the other parameters.

Impact: Virtual magnetic resonance elastography and intravoxel incoherent motion can quantify placental elasticity and circulation at the microscopic level, and are superior to the ultrasound parameters, in gestational hypertension. They may serve as a vital noninvasive supplement to assess placental dysfunction.

Keywords: Placenta, Quantitative Imaging, R2* mapping, Biomarkers, Contrast Agents, Ferumoxytol

Motivation: Pregnancy complications are often associated with placenta dysfunction, which lacks non-invasive assessment in early pregnancy.

Goal(s): Evaluating placental R2* as a non-invasive and quantitative biomarker to identify and monitor pregnancy complications.

Approach: We introduce a novel thrombotic and inflammatory placental injury model of rhesus macaques and report mean values, histograms, and spatial distribution of placental R2* based on blood oxygenation level dependent (BOLD) MR imaging at different gestational ages before and after iron nanoparticle (ferumoxytol) administration.

Results: Higher overtime and post-contrast increase in mean values and more heterogeneous spatial patterns of R2* are observed in placentas with induced injury.

Impact: A refined nonhuman primate model for thrombosis and inflammation is introduced for investigating placental pathology. Larger R2* increases over time and immediately after ferumoxytol injections suggest the potential for R2* to be an indicator of pathology with or without contrast.

Keywords: Placenta, Placenta

Motivation: The intravoxel incoherent motion (IVIM) model can separately assess diffusion in tissue and perfusion in vasculature. However, anisotropic extensions to IVIM that model coherently orientated vasculature are complex and difficult to fit.

Goal(s): Enhance the IVIM model to account for macroscopic anisotropy in vascular structures, while minimizing the increase in model complexity.

Approach: We model perfusion and diffusion compartments using constrained tensors and estimate the tensor parameters via the spherical mean.   

Results: Our spherical mean anisotropic IVIM approach quantifies and maps anisotropy in perfusion and diffusion compartments and captures microstructural and microcirculatory alterations in the placenta during pregnancy. 

Impact: Existing anisotropic IVIM models are complex and clinically impractical. We demonstrate a spherical mean approach that simplifies the disentanglement of perfusion- and diffusion-related anisotropy. This can enable rapid quantification of biomarkers for detecting microcirculatory and microstructural changes in anisotropic tissue.

Keywords: Placenta, Placenta, Quantitative flow

Motivation: Both oxygenation and blood flow are key measures of placental function; if we could dynamically, and simultaneously measure T2* and flow we could investigate placental function over time. 

Goal(s): To Implement and test a gradient echo, quantitative flow sequence with EPI readout in the placenta.

Approach: 4 participants were scanned at 3T using the sequence, and the T2* and velocity maps produced were investigated. 

Results: The T2* weighted image retained sufficient contrast, and the quantitative flow was repeatable across subjects, measuring similar velocities in the placenta (<0.1cm/s) that has previously been measured with MRI.

Impact: This sequence has promise for dynamically investigating placental function, simultaneously providing information about oxygenation and blood flow. This will allow us to evaluate placental function in compromised pregnancies over time, and investigate the role of placental contractions. 

Keywords: Pelvis, Diffusion/other diffusion imaging techniques, restriction spectrum imaging, cervical cancer

Motivation: Restriction spectrum imaging (RSI) can potentially improve cervical cancer evaluation pre- and post-therapy, particularly in separating persistent malignancy from post-treatment edema. 

Goal(s): To evaluate a cervical cancer-specific RSI model for assessment on an independent pilot cohort examined with combined PET/MRI.

Approach: Four patients underwent PET/MRI before (chemo)radiotherapy. Two of them also underwent PET/MRI within 4 weeks of treatment, MRI 3 months post-treatment and PET/MRI 6 months post-treatment. Two healthy volunteers underwent MRI only.

Results: Preliminary results showed increased tumor conspicuity on RSI compared to conventional pre-therapy PET/MRI and promising results early post-therapy.

Impact: Cervical cancer-specific restriction spectrum imaging (RSI), evaluated on an independent pilot cohort, indicates increased tumor conspicuity compared to conventional pre-therapy PET/MRI and promising results post-therapy. Future work will target early therapy response assessment in combined RSI-PET/MRI at a larger scale.

Keywords: Pelvis, Pelvis

Motivation: Hypoxia inducible factor (HIF-1α) is a major transcriptional factor regulating gene expression under hypoxic conditions and could serve as an important biomarker for tumor aggressiveness or radiation resistance.

Goal(s): This study aimed to investigate  whether the multimodal functional MRI technique can be used for quantitatively measuring HIF-1α expression.

Approach: APT, ADC, D, D^*, f, MK, and MD values were calculated and compared between HIF-1α high expression and HIF-1α low expression groups.

Results: APT, ADC, D, D*, MK and MD values were significantly higher in high HIF-1α expression than in low HIF-1α expression groups, whereas f value was significantly lower in high HIF-1α expression .

Impact: The quantitative parameters of APTw combined with multi-model diffusion-weighted sequences allowed quantitative assessment of EC HIF-1α expression, and the combined quantitative parameters further enhanced the assessment efficacy.

Keywords: Uterus, Uterus, Radiogenomics

Motivation: To explore genetically based molecular profiling of endometrial cancer (EC) patients to delineate prognostic risk groups.

Goal(s): To demonstrate the potential of radiogenomics for classification of EC molecular subtypes, 254 EC patients with histologically and genetically proven EC from two-center were enrolled.

Approach: A radiomics model based on four sequences was combined with genomics features to form the final diagnosed model.

Results: Our results showed a medium-to-high diagnostic performance to distinguish molecular subtypes with AUC of 0.849 and 0.673 in internal and external test sets, respectively. The radiogenomics model could guide clinicians in administering individual treatments for EC patients.

Impact: Our results demonstrate that the predictive model derived from MRI imaging features holds significant promise in identifying molecular subtypes in endometrial cancer. This model has the potential to guide clinicians in tailoring individualized treatments for EC patients.