Search

Domain adaptation for semantic segmentation requires pixel-level knowledge transfer from a labeled source domain to an unlabeled target domain. Existing approaches typically align the features of the source and target domains at different levels. However, they usually neglect the different adaptive complexities of different information flows within images. In this paper, we focus on combining two main information flows in semantic segmentation, ie., the pixel-level disparate information and image structure information. Specifically, we propose to combine two feature map-based prediction heads, which are thought to focus on pixel-level and structure-level information, to accommodate different complexities by adjusting the attention to adaptation functions of the target domain. We then align the outputs from the two heads through a consistency regularization to realize informative complementarity. The combined prediction head further enables regularizing the distance between different pixel representations of different classes, thereby mitigating the mis-adaptation problem of similar classes. The proposed method can achieve more competitive results than current state-of-the-art results on two publicly available benchmark datasets, ie., SYNTHIA → Cityscapes and GTA5 → Cityscapes. [ABSTRACT FROM AUTHOR]

B-cell activating factor (BAFF) and BAFF-receptor (BAFF-R) play crucial roles in the progression of malignant B-cells. The aim of the present study was to evaluate the expression profiles and the clinical significance of BAFF and BAFF-R in diffuse large B-cell lymphoma (DLBCL). Paraffin-embedded specimens from 136 patients with newly diagnosed DLBCL, treated with rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone chemotherapy (R-CHOP), were examined for BAFF and BAFF-R expression by immunohistochemistry. BAFF and BAFF-R were expressed in 72.1 % (98/136) and 47.1 % (64/136) of the DLBCL tissues, respectively. Negative BAFF-R expression was significantly correlated with elevated serum lactate dehydrogenase (LDH) levels (P = 0.036), an International Prognostic Index (IPI) score of 2 or higher (P < 0.001), and a poor revised IPI (R-IPI) risk score (P = 0.043). The complete response rate after R-CHOP was higher in patients with positive BAFF-R expression than in those with negative BAFF-R expression (73.4 vs. 56.9 %, P = 0.045). Negative expression of BAFF-R, but not of BAFF, was significantly associated with inferior progression-free survival (PFS; P = 0.020) and overall survival (OS; P = 0.028). Only negative BAFF-R expression was correlated with inferior PFS and OS in multivariate analysis (P = 0.049 and 0.040, respectively). Taken together, our results showed that the majority and approximate one-half of patients with DLBCL were positive for BAFF and BAFF-R, respectively. Negative expression of BAFF-R, but not of BAFF, could be an independent risk factor for PFS and OS in patients with DLBCL treated with standard R-CHOP. [ABSTRACT FROM AUTHOR]