Dense biased networks with deep priori anatomy and hard region adaptation: Semi-supervised learning for fine renal artery segmentation

Yuting He (1) , Guanyu Yang (1, 2) , Jian Yang (3) , Yang Chen (2, 1) , Youyong Kong (2, 1) , Jiasong Wu (1, 2) , Lijun Tang (4) , Xiaomei Zhu (4) , Jean-Louis Dillenseger (2, 5) , Pengfei Shao (4) , Shaobo Zhang (4) , Huazhong Shu (2, 1) , Jean-Louis Coatrieux (2, 5) , Shuo Li (6)

1 LIST - Laboratory of Image Science and Technology [Nanjing]
2 CRIBS - Centre de Recherche en Information Biomédicale sino-français
3 BIT - Beijing Institute of Technology
4 Nanjing Medical University
5 LTSI - Laboratoire Traitement du Signal et de l'Image
6 UWO - University of Western Ontario

Fine renal artery segmentation on abdominal CT angiography (CTA) image is one of the most important tasks for kidney disease diagnosis and pre-operative planning. It will help clinicians locate each interlobar artery's blood-feeding region via providing the complete 3D renal artery tree masks. However, it is still a task of great challenges due to the large intra-scale changes, large inter-anatomy variation, thin structures, small volume ratio and small labeled dataset of the fine renal artery. In this paper, we propose the first semi-supervised 3D fine re-nal artery segmentation framework, DPA-DenseBiasNet, which combines deep prior anatomy (DPA), dense biased network (DenseBiasNet) and hard region adaptation loss (HRA): 1) Based on our proposed dense biased connection, the DenseBiasNet fuses multi-receptive field and multi-resolution feature maps for large intra-scale changes. This dense biased connection also obtains a dense * Corresponding author information flow and dense gradient flow so that the training is accelerated and the accuracy is enhanced. 2) DPA features extracted from an autoencoder (AE) are embedded in DenseBiasNet to cope with the challenge of large inter-anatomy variation and thin structures. The AE is pre-trained (unsupervised) by numerous unlabeled data to achieve the representation ability of anatomy features and these features are embedded in DenseBiasNet. This process will not introduce incorrect labels as optimization targets and thus contributes to a stable semi-supervised training strategy that is suitable for sensitive thin structures. 3) The HRA selects the loss value calculation region dynamically according to the segmentation quality so the network will pay attention to the hard regions in the training process and keep the class balanced. Experiments demonstrated that DPA-DenseBiasNet had high predictive accuracy and generalization with the Dice coefficient of 0.884 which increased by 0.083 compared with 3D U-Net (Ç içek et al., 2016). This revealed our framework with great potential for the 3D fine renal artery segmentation in clinical practice.

Mots clés: dense biased connection; dense biased network; semi-supervised learning; renal artery segmentation; deep priori anatomy; CT angiography image; 3D fine segmentation; hard region adaptation loss function

Domaines: Imagerie; Imagerie médicale; Traitement du signal et de l'image [eess.SP]

Yuting He, Guanyu Yang, Jian Yang, Yang Chen, Youyong Kong, et al.. Dense biased networks with deep priori anatomy and hard region adaptation: Semi-supervised learning for fine renal artery segmentation. Medical Image Analysis, 2020, 63, pp.101722. ⟨10.1016/j.media.2020.101722⟩. ⟨hal-02614243⟩