Unsupervised domain adaptation aims to learn a model of classifier for unlabeled samples on the target domain, given training data of labeled samples on the source domain. Impressive progress is made recently by learning invariant features via domain-adversarial training of deep networks. In spite of the recent progress, domain adaptation is still limited in achieving the invariance of feature distributions at a finer category level. To this end, we propose in this paper a new domain adaptation method called Domain-Symmetric Networks (SymNets). The proposed SymNet is based on a symmetric design of source and target task classifiers, based on which we also construct an additional classifier that shares with them its layer neurons. To train the SymNet, we propose a novel adversarial learning objective whose key design is based on a two-level domain confusion scheme, where the category-level confusion loss improves over the domain-level one by driving the learning of intermediate network features to be invariant at the corresponding categories of the two domains. Both domain discrimination and domain confusion are implemented based on the constructed additional classifier. Since target samples are unlabeled, we also propose a scheme of cross-domain training to help learn the target classifier. Careful ablation studies show the efficacy of our proposed method. In particular, based on commonly used base networks, our SymNets achieve the new state of the art on three benchmark domain adaptation datasets.

With the introduction of fully convolutional neural networks, deep learning has raised the benchmark for medical image segmentation on both speed and accuracy, and different networks have been proposed for 2D and 3D segmentation with promising results. Nevertheless, most networks only handle relatively small numbers of labels (<10), and there are very limited works on handling highly unbalanced object sizes especially in 3D segmentation. In this paper, we propose a network architecture and the corresponding loss function which improve segmentation of very small structures. By combining skip connections and deep supervision with respect to the computational feasibility of 3D segmentation, we propose a fast converging and computationally efficient network architecture for accurate segmentation. Furthermore, inspired by the concept of focal loss, we propose an exponential logarithmic loss which balances the labels not only by their relative sizes but also by their segmentation difficulties. We achieve an average Dice coefficient of 82% on brain segmentation with 20 labels, with the ratio of the smallest to largest object sizes as 0.14%. Less than 100 epochs are required to reach such accuracy, and segmenting a 128×128×128 volume only takes around 0.4 s.

Unsupervised domain adaptation (UDA) is to learn classification models that make predictions for unlabeled data on a target domain, given labeled data on a source domain whose distribution diverges from the target one. Mainstream UDA methods strive to learn domain-aligned features such that classifiers trained on the source features can be readily applied to the target ones. Although impressive results have been achieved, these methods have a potential risk of damaging the intrinsic data structures of target discrimination, raising an issue of generalization particularly for UDA tasks in an inductive setting. To address this issue, we are motivated by a UDA assumption of structural similarity across domains, and propose to directly uncover the intrinsic target discrimination via constrained clustering, where we constrain the clustering solutions using structural source regularization that hinges on the very same assumption. Technically, we propose a hybrid model of Structurally Regularized Deep Clustering, which integrates the regularized discriminative clustering of target data with a generative one, and we thus term our method as SRDC++. Our hybrid model is based on a deep clustering framework that minimizes the Kullback-Leibler divergence between the distribution of network prediction and an auxiliary one, where we impose structural regularization by learning domain-shared classifier and cluster centroids. By enriching the structural similarity assumption, we are able to extend SRDC++ for a pixel-level UDA task of semantic segmentation. We conduct extensive experiments on seven UDA benchmarks of image classification and semantic segmentation. With no explicit feature alignment, our proposed SRDC++ outperforms all the existing methods under both the inductive and transductive settings. We make our implementation codes publicly available at https://github.com/huitangtang/SRDCPP.

New translational strategies are needed to improve diabetes outcomes among low-income African-Americans. Our goal was to develop/pilot test a patient intervention combining culturally tailored diabetes education with shared decision-making training. This was an observational cohort study. Surveys and clinical data were collected at baseline, program completion, and 3 and 6 months. There were 21 participants; the mean age was 61 years. Eighty-six percent of participants attended >70 % of classes. There were improvements in diabetes self-efficacy, self-care behaviors (i.e., following a "healthful eating plan" (mean score at baseline 3.4 vs. 5.2 at program's end; p = 0.002), self glucose monitoring (mean score at baseline 4.3 vs. 6.2 at program's end; p = 0.04), and foot care (mean score at baseline 4.1 vs. 6.0 at program's end; p = 0.001)), hemoglobin A1c (8.24 at baseline vs. 7.33 at 3-month follow-up, p = 0.02), and HDL cholesterol (51.2 at baseline vs. 61.8 at 6-month follow-up, p = 0.01). Combining tailored education with shared decision-making may be a promising strategy for empowering low-income African-Americans and improving health outcomes.

Background We sought to determine whether perceived patient-centered medical home (PCMH) characteristics are associated with staff morale, job satisfaction, and burnout in safety net clinics. Methods Self-administered survey among 391 providers and 382 clinical staff across 65 safety net clinics in 5 states in 2010. The following 5 subscales measured respondents’ perceptions of PCMH characteristics on a scale of 0 to 100 (0 indicates worst and 100 indicates best): access to care and communication with patients, communication with other providers, tracking data, care management, and quality improvement. The PCMH sub-scale scores were averaged to create a total PCMH score. Results Six hundred three persons (78.0%) responded. In multivariate generalized estimating equation models, a 10% increase in the quality improvement subscale score was associated with higher morale (provider odds ratio [OR], 2.64; 95% CI, 1.47–4.75; staff OR, 3.62; 95% CI, 1.84–7.09), greater job satisfaction (provider OR, 2.45; 95% CI, 1.42–4.23; staff OR, 2.55; 95% CI 1.42–4.57), and freedom from burnout (staff OR, 2.32; 95% CI, 1.31–4.12). The total PCMH score was associated with higher staff morale (OR, 2.63; 95% CI, 1.47–4.71) and with lower provider freedom from burnout (OR, 0.48; 95% CI, 0.30–0.77). A separate work environment covariate correlated highly with the quality improvement subscale score and the total PCMH score, and PCMH characteristics had attenuated associations with morale and job satisfaction when included in models. Conclusions Providers and staff who perceived more PCMH characteristics in their clinics were more likely to have higher morale, but the providers had less freedom from burnout. Among the PCMH subscales, the quality improvement subscale score particularly correlated with higher morale, greater job satisfaction, and freedom from burnout.