Endometrial cancer (EC) has the highest rate of malignancy in women in the entire world, including China, which has the largest population. Accurately staging EC prior to an invasive procedure still poses a challenge for clinicians. In the present study, we used more than five hundred EC patients’ MR images to train the computer to establish a deep learning diagnostic model to help the doctor automatically identify the myometrial invasion. We trained a detection model based on the YOLOv3 algorithm to locate the lesion area on EC MR Imaging. Subsequently, the detected regions were fed into a classification model based on a deep learning network to automatically identify myometrial invasion depth. Together, the radiologists and trained network model yielded an accuracy of 86.2% in determining deep myometrial invasion. In this study, the deep learning network model derived from EC MR Imaging provided a competitive, time-efficient diagnostic performance in myometrial invasion depth identification.
- The models established with the deep learning method could help improve the diagnostic confidence and performance of MI identification based on endometrial cancer MR imaging.
- The models enabled the classification of endometrial cancer MR images to the two categories with a sensitivity of 0.67, a specificity of 0.88, and an accuracy of 0.85.
- Using the detected lesion region to evaluate myometrial invasion depth could remove redundant information in the image and provide more effective features.
Article: Deep learning for the determination of myometrial invasion depth and automatic lesion identification in endometrial cancer MR imaging: a preliminary study in a single institution
Authors: Xiaojun Chen, Yida Wang, Minhua Shen, Bingyi Yang, Qing Zhou, Yinqiao Yi, Weifeng Liu, Guofu Zhang, Guang Yang & He Zhang