Transesophageal echocardiogram (TEE) can promote the quality of cardiac surgery and reduce peri-operative complications, and thus has been gradually accepted by cardiac surgeons. Through an esophageal probe, TEE can clearly visualize the internal structure of the heart without interrupting surgical procedure. As a newly developed technology which breaks the limitations tied to the traditional two-dimensional TEE, the realtime threedimensional transesophageal echocardiogram (RT3D-TEE) has the advantages of showing threedimensional structure of the heart and providing full range of anatomical information of the heart. Furthermore, it can precisely analyze the anatomical structure of the abnormal heart valves and provide assessment of the change of heart volume. Relying on its unique imaging property, it can largely facilitate preoperative decision-making and provide realtime intraoperative guidance as well as accurate postoperative evaluation. It has now been successfully applied in various types of cardiac surgical procedures including valve repair surgery, congenital heart defect intervention, cardiac mass removal as well as heart function evaluation. In this article, we will review the applications of RT3D-TEE in cardiac surgery, and try to form a basis for its further clinical application.
This paper performs a comprehensive study on the computer-aided detection for the medical diagnosis with deep learning. Based on the region convolution neural network and the prior knowledge of target, this algorithm uses the region proposal network, the region of interest pooling strategy, introduces the multi-task loss function: classification loss, bounding box localization loss and object rotation loss, and optimizes it by end-to-end. For medical image it locates the target automatically, and provides the localization result for the next stage task of segmentation. For the detection of left ventricular in echocardiography, proposed additional landmarks such as mitral annulus, endocardial pad and apical position, were used to estimate the left ventricular posture effectively. In order to verify the robustness and effectiveness of the algorithm, the experimental data of ultrasonic and nuclear magnetic resonance images are selected. Experimental results show that the algorithm is fast, accurate and effective.