Protect the Brain When Treating the Heart: A Convolutional Neural Network for Detecting Emboli
For clinicians performing cardiac interventions, this provides a real-time, automated tool to detect and quantify microemboli, a common complication, potentially improving patient safety.
The authors developed a 2.5D U-Net to segment gaseous microemboli in cardiac ultrasound, achieving robust detection and high segmentation accuracy with real-time speed, enabling integration into surgical monitoring for quantification of emboli area over time.
Gaseous microemboli (GME) represent a common complication of cardiac structural interventions across both surgical and transcatheter approaches. Transthoracic cardiac ultrasound imaging represents a convenient methodology to visualize the presence of circulating GME. However, their detection and quantification are far from trivial due to operator-dependent view, high velocity, and objects with similar structure in the background. Here, we propose an approach based on a 2.5D U-Net architecture to segment GME in space-time connected data. Such an approach yields robust detection against the background and high segmentation accuracy while retaining real-time execution speed. These properties facilitated the integration of the proposed pipeline into patient-monitoring surgical protocols, providing the quantification of GME area over time.