Abstract:
A new method is proposed to reconstruct and analyse the left ventricle (LV) from multiple acoustic windows 3D ultrasound acquired using a 3D rotational probe. Prior research in this area has been based on one acoustic window acquisition. However, the data suffers from several limitations that degrade the 3D reconstruction, such as motion of the probe during the acquisition and the presence of shadow due to bone (ribs) and air (in the lungs). In this paper we aim to overcome these limitations by automatically fusing information from multiple acoustic windows sparse-view acquisitions and using a position sensor to track the probe in real time. Geometric constraints of the object shape, and spatio-temporal information relating to the image acquisition process are used to propose new algorithms for (1) grouping endocardial edge cues from an initial image segmentation and (2) defining a novel reconstruction method that utilises information from multiple acoustic windows. We illustrate our new method on phantom and real heart data and compare its performance against our previous approach that is based on a single acoustic window.
