LE-UDA: Label-Efficient Unsupervised Domain Adaptation for Medical Image Segmentation.

While deep learning methods hitherto have achieved considerable success in medical image segmentation, they are still hampered by two limitations: (i) reliance on large-scale well-labeled datasets, which are difficult to curate due to the expert-driven and time-consuming nature of pixel-level annotations in clinical practices, and (ii) failure to generalize from one domain to another, especially when the target domain is a different modality with severe domain shifts. Recent unsupervised domain adaptation (UDA) techniques leverage abundant labeled source data together with unlabeled target data to reduce the domain gap, but these methods degrade significantly with limited source annotations. In this study, we address this underexplored UDA problem, investigating a challenging but valuable realistic scenario, where the source domain not only exhibits domain shift w.r.t. the target domain but also suffers from label scarcity. In this regard, we propose a novel and generic framework called "Label-Efficient Unsupervised Domain Adaptation" (LE-UDA). In LE-UDA, we construct self-ensembling consistency for knowledge transfer between both domains, as well as a self-ensembling adversarial learning module to achieve better feature alignment for UDA. To assess the effectiveness of our method, we conduct extensive experiments on two different tasks for cross-modality segmentation between MRI and CT images. Experimental results demonstrate that the proposed LE-UDA can efficiently leverage limited source labels to improve cross-domain segmentation performance, outperforming state-of-the-art UDA approaches in the literature.

[Evaluation of comprehensive measures of schistosomiasis control in Caidian District, Wuhan City].

OBJECTIVE: To evaluate the implementation and effectiveness of comprehensive control measures of schistosomiasis in Caidian District, Wuhan City. METHODS: The data of implementation of the schistosomiasis control measures in Caidian District was collected and analyzed statistically. RESULTS: The number of Oncomelania hupensis snail environments and snail area decreased from 213 sites and 223.47 hm2 in 2003 to 59 sites and 51.20 hm2 in 2013, with the decreasing rate of 72.30% and 77.09%, respectively. The infection rate of schistosomiasis of human decreased from 1.58% in 2006 to 0.24% in 2013, with the decreasing rate of 84.81%. The positive rate of stool tests of cattle was 2.19% in 2006 while it was 0 in 2013. CONCLUSION: The comprehensive control measures of schistosomiasis are effective.

[Estimation of vegetation canopy water content using Hyperion hyperspectral data].

Vegetation canopy water content (VCWC) has widespread utility in agriculture, ecology and hydrology. Based on the PROSAIL model, a novel model for quantitative inversion of vegetation canopy water content using Hyperion hyperspectral data was explored. Firstly, characteristics of vegetation canopy reflection were investigated with the PROSAIL radiative transfer model, and it was showed that the first derivative at the right slope (980 - 1 070 nm) of the 970 nm water absorption feature (D98-1 070) was closely related to VCWC, and determination coefficient reached to 0.96. Then, bands 983, 993, 1 003, 1 013, 1 023, 1 033, 1 043, 1 053 and 1 063 nm of Hyperion data were selected to calculate D980-1 070, and VCWC was estimated using the proposed method. Finally, the retrieval result was verified using field measured data in Yingke oasis of the Heihe basin. It indicated that the mean relative error was 12.5%, RMSE was within 0.1 kg x m(-2) and the proposed model was practical and reliable. This study provides a more efficient way for obtaining VCWC of large area.