Noise-Robust Sleep Staging via Adversarial Training With an Auxiliary Model.

Deep learning (DL)-based automatic sleep staging approaches have attracted much attention recently due in part to their outstanding accuracy. At the testing stage, however, the performance of these approaches is likely to be degraded, when applied in different testing environments, because of the problem of domain shift. This is because while a pre-trained model is typically trained on noise-free electroencephalogram (EEG) signals acquired from accurate medical equipment, deployment is carried out on consumer-level devices with undesirable noise. To alleviate this challenge, in this work, we propose an efficient training approach that is robust against unseen arbitrary noise. In particular, we propose to generate the worst-case input perturbations by means of adversarial transformation in an auxiliary model, to learn a wide range of input perturbations and thereby to improve reliability. Our approach is based on two separate training models: (i) an auxiliary model to generate adversarial noise and (ii) a target network to incorporate the noise signal to enhance robustness. Furthermore, we exploit novel class-wise robustness during the training of the target network to represent different robustness patterns of each sleep stage. Our experimental results demonstrated that our approach improved sleep staging performance on healthy controls, in the presence of moderate to severe noise levels, compared with competing methods. Our approach was able to effectively train and deploy a DL model to handle different types of noise, including adversarial, Gaussian, and shot noise.

Unsupervised Domain Adaptation for Segmentation with Black-box Source Model.

Unsupervised domain adaptation (UDA) has been widely used to transfer knowledge from a labeled source domain to an unlabeled target domain to counter the difficulty of labeling in a new domain. The training of conventional solutions usually relies on the existence of both source and target domain data. However, privacy of the large-scale and well-labeled data in the source domain and trained model parameters can become the major concern of cross center/domain collaborations. In this work, to address this, we propose a practical solution to UDA for segmentation with a black-box segmentation model trained in the source domain only, rather than original source data or a white-box source model. Specifically, we resort to a knowledge distillation scheme with exponential mixup decay (EMD) to gradually learn target-specific representations. In addition, unsupervised entropy minimization is further applied to regularization of the target domain confidence. We evaluated our framework on the BraTS 2018 database, achieving performance on par with white-box source model adaptation approaches.

Unsupervised Black-Box Model Domain Adaptation for Brain Tumor Segmentation.

Unsupervised domain adaptation (UDA) is an emerging technique that enables the transfer of domain knowledge learned from a labeled source domain to unlabeled target domains, providing a way of coping with the difficulty of labeling in new domains. The majority of prior work has relied on both source and target domain data for adaptation. However, because of privacy concerns about potential leaks in sensitive information contained in patient data, it is often challenging to share the data and labels in the source domain and trained model parameters in cross-center collaborations. To address this issue, we propose a practical framework for UDA with a black-box segmentation model trained in the source domain only, without relying on source data or a white-box source model in which the network parameters are accessible. In particular, we propose a knowledge distillation scheme to gradually learn target-specific representations. Additionally, we regularize the confidence of the labels in the target domain via unsupervised entropy minimization, leading to performance gain over UDA without entropy minimization. We extensively validated our framework on a few datasets and deep learning backbones, demonstrating the potential for our framework to be applied in challenging yet realistic clinical settings.

Transferring Structured Knowledge in Unsupervised Domain Adaptation of a Sleep Staging Network.

Automatic sleep staging based on deep learning (DL) has been attracting attention for analyzing sleep quality and determining treatment effects. It is challenging to acquire long-term sleep data from numerous subjects and manually labeling them even though most DL-based models are trained using large-scale sleep data to provide state-of-the-art performance. One way to overcome this data shortage is to create a pre-trained network with an existing large-scale dataset (source domain) that is applicable to small cohorts of datasets (target domain); however, discrepancies in data distribution between the domains prevent successful refinement of this approach. In this paper, we propose an unsupervised domain adaptation method for sleep staging networks to reduce discrepancies by re-aligning the domains in the same space and producing domain-invariant features. Specifically, in addition to a classical domain discriminator, we introduce local discriminators - subject and stage - to maintain the intrinsic structure of sleep data to decrease local misalignments while using adversarial learning to play a minimax game between the feature extractor and discriminators. Moreover, we present several optimization schemes during training because the conventional adversarial learning is not effective to our training scheme. We evaluate the performance of the proposed method by examining the staging performances of a baseline network compared with direct transfer (DT) learning in various conditions. The experimental results demonstrate that the proposed domain adaptation significantly improves the performance though it needs no labeled sleep data in target domain.

Interferon ß protects against lethal endotoxic and septic shock through SIRT1 upregulation.

Lipopolysaccharide (LPS), an endotoxin derived from gram-negative bacteria, promotes the secretion of proinflammatory cytokines and mediates endotoxemia through activation of mitogen activated protein kinases, NF-κB, and interferon regulatory factor-3. Silent information regulator transcript-1 (SIRT1), an NAD-dependent deacetylase, mediates NF-κB deacetylation, and inhibits its function. SIRT1 may affect LPS-mediated signaling pathways and endotoxemia. Here we demonstrate that SIRT1 blocks LPS-induced secretion of interleukin 6 and tumor necrosis factor α in murine macrophages, and protects against lethal endotoxic and septic shock in mice. We also demonstrate that interferon ß increases SIRT1 expression by activating the Janus kinase--signal transducer and activator of transcription (JAK-STAT) pathway in mouse bone marrow derived macrophages. In vivo treatment of interferon ß protects against lethal endotoxic and septic shock, which is abrogated by infection with dominant negative SIRT1-expressing adenovirus. Our work suggests that both SIRT1 and SIRT1-inducing cytokines are useful targets for treating patients with sepsis.

Lipopolysaccharide induces CD38 expression and solubilization in J774 macrophage cells.

CD38, an ADP ribosyl cyclase, is a 45 kDa type II transmembrane protein having a short N-terminal cytoplasmic domain and a long C-terminal extracellular domain, expressed on the surface of various cells including macrophages, lymphocytes, and pancreatic ß cells. It is known to be involved in cell adhesion, signal transduction and calcium signaling. In addition to its transmembrane form, CD38 is detectable in biological fluids in soluble forms. The mechanism by which CD38 is solubilized from the plasma membrane is not yet clarified. In this study, we found that lipopolysaccharide (LPS) induced CD38 upregulation and its extracellular release in J774 macrophage cells. Furthermore, it also increased CD38 expression at the mRNA level by activating the Janus kinase-signal transducers and activators of transcription (JAK-STAT) pathway. However, LPS decreased the levels of CD38 in the plasma membrane by releasing CD38 into the culture supernatant. LPS-induced CD38 release was blocked by the metalloproteinase-9 inhibitor indicating that MMP-9 solubilizes CD38. In conclusion, the present findings demonstrate a potential mechanism by which C38 is solubilized from the plasma membrane.

NAADP mediates insulin-stimulated glucose uptake and insulin sensitization by PPARγ in adipocytes.

Insulin stimulates glucose uptake through the membrane translocation of GLUT4 and GLUT1. Peroxisome proliferator-activated receptor γ (PPARγ) enhances insulin sensitivity. Here, we demonstrate that insulin stimulates GLUT4 and GLUT1 translocation, and glucose uptake, by activating the signaling pathway involving nicotinic acid adenine dinucleotide phosphate (NAADP), a calcium mobilizer, in adipocytes. We also demonstrate that PPARγ mediates insulin sensitization by enhancing NAADP production through upregulation of CD38, the only enzyme identified for NAADP synthesis. Insulin produced NAADP by both CD38-dependent and -independent pathways, whereas PPARγ produced NAADP by CD38-dependent pathway. Blocking the NAADP signaling pathway abrogated both insulin-stimulated and PPARγ-induced GLUT4 and GLUT1 translocation, thereby inhibiting glucose uptake. CD38 knockout partially inhibited insulin-stimulated glucose uptake. However, CD38 knockout completely blocked PPARγ-induced glucose uptake in adipocytes and PPARγ-mediated amelioration of glucose tolerance in diabetic mice. These results demonstrated that the NAADP signaling pathway is a critical molecular target for PPARγ-mediated insulin sensitization.

Connexin-43 hemichannels mediate cyclic ADP-ribose generation and its Ca2+-mobilizing activity by NAD+/cyclic ADP-ribose transport.

The ADP-ribosyl cyclase CD38 whose catalytic domain resides in outside of the cell surface produces the second messenger cyclic ADP-ribose (cADPR) from NAD(+). cADPR increases intracellular Ca(2+) through the intracellular ryanodine receptor/Ca(2+) release channel (RyR). It has been known that intracellular NAD(+) approaches ecto-CD38 via its export by connexin (Cx43) hemichannels, a component of gap junctions. However, it is unclear how cADPR extracellularly generated by ecto-CD38 approaches intracellular RyR although CD38 itself or nucleoside transporter has been proposed to import cADPR. Moreover, it has been unknown what physiological stimulation can trigger Cx43-mediated export of NAD(+). Here we demonstrate that Cx43 hemichannels, but not CD38, import cADPR to increase intracellular calcium through RyR. We also demonstrate that physiological stimulation such as Fcγ receptor (FcγR) ligation induces calcium mobilization through three sequential steps, Cx43-mediated NAD(+) export, CD38-mediated generation of cADPR and Cx43-mediated cADPR import in J774 cells. Protein kinase A (PKA) activation also induced calcium mobilization in the same way as FcγR stimulation. FcγR stimulation-induced calcium mobilization was blocked by PKA inhibition, indicating that PKA is a linker between FcγR stimulation and NAD(+)/cADPR transport. Cx43 knockdown blocked extracellular cADPR import and extracellular cADPR-induced calcium mobilization in J774 cells. Cx43 overexpression in Cx43-negative cells conferred extracellular cADPR-induced calcium mobilization by the mediation of cADPR import. Our data suggest that Cx43 has a dual function exporting NAD(+) and importing cADPR into the cell to activate intracellular calcium mobilization.