Hospital patients can have catheters and lines inserted during their admission to give medicines or to do medical tests quickly, especially central venous catheters (CVCs). Following the insertion process, the position of the catheter is commonly verified based on the X-ray images to detect the malposition, avoiding a series of complications. Recently, deep learning frameworks have shown their potential to assist in detecting the malposition of the CVC in X-ray images. However, these deep learning approaches meet three main challenges: 1) The absence and ambiguous anatomical landmarks in X-ray images make the implementation of the landmark-annotation-based framework particularly challenging. 2) Most of the existing deep learning frameworks have a large number of parameters and a large computation workload. 3) The feature significance for the deep-learning framework to detect the misplaced catheter is not well studied, which may impact the understanding of the deep-learning model results. Therefore, a framework based on a graph convolutional neural network (GCN) is proposed in this work to detect catheter malposition. It achieves an AUC of 0.867 with 0.0229 million parameters. Even with a landmark missing, it still achieves an AUC of 0.855. It is demonstrated that the proposed framework can tolerate the missing and ambiguous landmarks in chest X-rays (CXRs) and get acceptable detection performance. Additionally, the proposed method achieves state-of-the-art performance with a relatively small number of parameters. Moreover, the features much more important for catheter malposition detection can be verified by the proposed framework to make the catheter malposition detection framework more understandable.

Hospital patients can have catheters and lines inserted during the course of their admission to give medicines for the treatment of medical issues, especially the central venous catheter (CVC). However, malposition of CVC will lead to many complications, even death. Clinicians always detect the malposition based on position detection of CVC tip via X-ray images. To reduce the workload of the clinicians and the percentage of malposition occurrence, we propose an automatic catheter tip detection framework based on a convolutional neural network (CNN). The proposed framework contains three essential components which are modified HRNet, segmentation supervision module, and deconvolution module. The modified HRNet can retain high-resolution features from start to end, ensuring the maintenance of precise information from the X-ray images. The segmentation supervision module can alleviate the presence of other line-like structures such as the skeleton as well as other tubes and catheters used for treatment. In addition, the deconvolution module can further increase the feature resolution on the top of the highest-resolution feature maps in the modified HRNet to get a higher-resolution heatmap of the catheter tip. A public CVC Dataset is utilized to evaluate the performance of the proposed framework. The results show that the proposed algorithm offering a mean Pixel Error of 4.11 outperforms three comparative methods (Ma's method, SRPE method, and LCM method). It is demonstrated to be a promising solution to precisely detect the tip position of the catheter in X-ray images.

Surgical instrument segmentation is crucial for computer-assisted surgery. Different from common object segmentation, it is more challenging due to the large illumination variation and scale variation in the surgical scenes. In this paper, we propose a bilinear attention network with adaptive receptive fields to address these two issues. To deal with the illumination variation, the bilinear attention module models global contexts and semantic dependencies between pixels by capturing second-order statistics. With them, semantic features in challenging areas can be inferred from their neighbors, and the distinction of various semantics can be boosted. To adapt to the scale variation, our adaptive receptive field module aggregates multi-scale features and selects receptive fields adaptively. Specifically, it models the semantic relationships between channels to choose feature maps with appropriate scales, changing the receptive field of subsequent convolutions. The proposed network achieves the best performance 97.47% mean IoU on Cata7. It also takes the first place on EndoVis 2017, exceeding the second place by 10.10% mean IoU.