Abstract
Automatic segmentation of gastric tumor not only provides image-guided clinical diagnosis but also assists radiologists to read images and improve the diagnostic accuracy. However, due to the inhomogeneous intensity distribution of gastric tumors in CT scans, the ambiguous/missing boundaries, and the highly variable shapes of gastric tumors, it is quite challenging to develop an automatic solution. This study designs a novel 3D improved feature pyramidal network (3D IFPN) to automatically segment gastric tumors in computed tomography (CT) images. To meet the challenges of this extremely difficult task, the proposed 3D IFPN makes full use of the complementary information within the low and high layers of deep convolutional neural networks, which is equipped with three types of feature enhancement modules: 3D adaptive spatial feature fusion (ASFF) module, single-level feature refinement (SLFR) module, and multi-level feature refinement (MLFR) module. The 3D ASFF module adaptively suppresses the feature inconsistency in different levels and hence obtains the multi-level features with high feature invariance. Then, the SLFR module combines the adaptive features and previous multi-level features at each level to generate the multi-level refined features by skip connection and attention mechanism. The MLFR module adaptively recalibrates the channel-wise and spatial-wise responses by adding the attention operation, which improves the prediction capability of the network. Furthermore, a stage-wise deep supervision (SDS) mechanism and a hybrid loss function are also embedded to enhance the feature learning ability of the network. CT volumes dataset collected in three Chinese medical centers was used to evaluate the segmentation performance of the proposed 3D IFPN model. Experimental results indicate that our method outperforms state-of-the-art segmentation networks in gastric tumor segmentation. Moreover, to explore the generalization for other segmentation tasks, we also extend the proposed network to liver tumor segmentation in CT images of the MICCAI 2017 Liver Tumor Segmentation Challenge.
Highlights
Gastric cancer, a very commonly diagnosed cancer of the digestive system, is the second leading cause of cancer death in China [1], which brings a heavy burden to the family and society
The proposed 3D improved feature pyramid network (3D IFPN) is evaluated on a self-collected computed tomography (CT) image dataset acquired from three Chinese medical centers, which achieves quite promising gastric tumor segmentation performance and outperforms other state-of-the-art methods
The DAF3D model is an improved version of feature pyramid network (FPN) with the equipment of attention modules refining deep attentive features at each layer, which depends on the complementary learning of both semantics and fine features at different levels
Summary
A very commonly diagnosed cancer of the digestive system, is the second leading cause of cancer death in China [1], which brings a heavy burden to the family and society. Accurate boundary detection and early staging of the neoplasm are favorable for surgical management optimization. As a convenient imaging examination tool, computed tomography (CT) can non-invasively provide the anatomical detail of the gastric tumor in a short time through the panoramic view. It is possible to recognize the single gastric wall layers as well as to estimate the invasive depth of the neoplasm on CT images [2], which is essential for tumor staging [3] and edge delineation. Accurate boundary is of great importance in volume assessment [4], further radiomics feature analysis [5] and image-guided navigation [6], the precise CT-based tumor segmentation is quite desirable. Thanks to the development of artificial intelligence, tumor segmentation can be done in a more automatic way
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call