Towards Automated Ultrasound Transesophageal Echocardiography and X-Ray Fluoroscopy Fusion Using an Image-Based Co-registration Method

Sun, Shanhui; Miao, Shun; Heimann, Tobias; Chen, Terrence; Kaiser, Markus; John, Matthias; Girard, Erin; Liao, Rui

doi:10.1007/978-3-319-46720-7_46

Cited by 2 publications

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Region-of-interest (ROI) extraction is performed via probabilistic boosting tree (PBT) detectors. 5 Following the convolutional layers and pooling layers, there is one FC layer with 1024 neurons, and the last FC layer then outputs the registration parameters. More details of the framework can be found in the preliminary version of this paper.…”

Section: -D/3-d Registration Problem Definitionmentioning

confidence: 99%

“…An optimizer is employed to maximize an intensity-based similarity measure between the DRR and x-ray images. 4,5 Although optimization-based methods are accurate, their computational efficiency is limited since they usually need many iterations of DRR generation and similarity computation. In addition, pose initialization in a close neighborhood of the correct pose is often required due to the small capture range.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Pairwise domain adaptation module for CNN-based 2-D/3-D registration

et al. 2018

Self Cite

View full text Add to dashboard Cite

Accurate two-dimensional to three-dimensional (2-D/3-D) registration of preoperative 3-D data and intraoperative 2-D x-ray images is a key enabler for image-guided therapy. Recent advances in 2-D/3-D registration formulate the problem as a learning-based approach and exploit the modeling power of convolutional neural networks (CNN) to significantly improve the accuracy and efficiency of 2-D/3-D registration. However, for surgery-related applications, collecting a large clinical dataset with accurate annotations for training can be very challenging or impractical. Therefore, deep learning-based 2-D/3-D registration methods are often trained with synthetically generated data, and a performance gap is often observed when testing the trained model on clinical data. We propose a pairwise domain adaptation (PDA) module to adapt the model trained on source domain (i.e., synthetic data) to target domain (i.e., clinical data) by learning domain invariant features with only a few paired real and synthetic data. The PDA module is designed to be flexible for different deep learning-based 2-D/3-D registration frameworks, and it can be plugged into any pretrained CNN model such as a simple Batch-Norm layer. The proposed PDA module has been quantitatively evaluated on two clinical applications using different frameworks of deep networks, demonstrating its significant advantages of generalizability and flexibility for 2-D/3-D medical image registration when a small number of paired real-synthetic data can be obtained.

show abstract

Section: -D/3-d Registration Problem Definitionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%