Predictive model for the preoperative assessment and prognostic modeling of lymph node metastasis in endometrial cancer.

Lymph node metastasis (LNM) is a well-established prognostic factor in endometrial cancer (EC). We aimed to construct a model that predicts LNM and prognosis using preoperative factors such as myometrial invasion (MI), enlarged lymph nodes (LNs), histological grade determined by endometrial biopsy, and serum cancer antigen 125 (CA125) level using two independent cohorts consisting of 254 EC patients. The area under the receiver operating characteristic curve (AUC) of the constructed model was 0.80 regardless of the machine learning techniques. Enlarged LNs and higher serum CA125 levels were more significant in patients with low-grade EC (LGEC) and LNM than in patients without LNM, whereas deep MI and higher CA125 levels were more significant in patients with high-grade EC (HGEC) and LNM than in patients without LNM. The predictive performance of LNM in the HGEC group was higher than that in the LGEC group (AUC = 0.84 and 0.75, respectively). Patients in the group without postoperative pathological LNM and positive LNM prediction had significantly worse relapse-free and overall survival than patients with negative LNM prediction (log-rank test, P < 0.01). This study showed that preoperative clinicopathological factors can predict LNM with high precision and detect patients with poor prognoses. Furthermore, clinicopathological factors associated with LNM were different between HGEC and LGEC patients.

Automated Endocardial Border Detection and Left Ventricular Functional Assessment in Echocardiography Using Deep Learning.

Endocardial border detection is a key step in assessing left ventricular systolic function in echocardiography. However, this process is still not sufficiently accurate, and manual retracing is often required, causing time-consuming and intra-/inter-observer variability in clinical practice. To address these clinical issues, more accurate and normalized automatic endocardial border detection would be valuable. Here, we develop a deep learning-based method for automated endocardial border detection and left ventricular functional assessment in two-dimensional echocardiographic videos. First, segmentation of the left ventricular cavity was performed in the six representative projections for a cardiac cycle. We employed four segmentation methods: U-Net, UNet++, UNet3+, and Deep Residual U-Net. UNet++ and UNet3+ showed a sufficiently high performance in the mean value of intersection over union and Dice coefficient. The accuracy of the four segmentation methods was then evaluated by calculating the mean value for the estimation error of the echocardiographic indexes. UNet++ was superior to the other segmentation methods, with the acceptable mean estimation error of the left ventricular ejection fraction of 10.8%, global longitudinal strain of 8.5%, and global circumferential strain of 5.8%, respectively. Our method using UNet++ demonstrated the best performance. This method may potentially support examiners and improve the workflow in echocardiography.