Alzheimer's disease (AD) is an age-related neurodegenerative disorder that displays pathological characteristics including senile plaques and neurofibrillary tangles. Metabolic defects are also present in AD-brain: for example, signs of deficient cerebral glucose uptake may occur decades before onset of cognitive dysfunction and tissue damage. There have been few systematic studies of the metabolite content of AD human brain, possibly due to scarcity of high-quality brain tissue and/or lack of reliable experimental methodologies. Here we sought to: 1) elucidate the molecular basis of metabolic defects in human AD-brain; and 2) identify endogenous metabolites that might guide new approaches for therapeutic intervention, diagnosis or monitoring of AD. Brains were obtained from nine cases with confirmed clinical/neuropathological AD and nine controls matched for age, sex and post-mortem delay. Metabolite levels were measured in post-mortem tissue from seven regions: three that undergo severe neuronal damage (hippocampus, entorhinal cortex and middle-temporal gyrus); three less severely affected (cingulate gyrus, sensory cortex and motor cortex); and one (cerebellum) that is relatively spared. We report a total of 55 metabolites that were altered in at least one AD-brain region, with different regions showing alterations in between 16 and 33 metabolites. Overall, we detected prominent global alterations in metabolites from several pathways involved in glucose clearance/utilization, the urea cycle, and amino-acid metabolism. The finding that potentially toxigenic molecular perturbations are widespread throughout all brain regions including the cerebellum is consistent with a global brain disease process rather than a localized effect of AD on regional brain metabolism.

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that currently affects 36 million people worldwide with no effective treatment available. Development of AD follows a distinctive pattern in the brain and is poorly modelled in animals. Therefore, it is vital to widen the spatial scope of the study of AD and prioritise the study of human brains. Here we show that functionally distinct human brain regions display varying and region-specific changes in protein expression. These changes provide insights into the progression of disease, novel AD-related pathways, the presence of a gradient of protein expression change from less to more affected regions and a possibly protective protein expression profile in the cerebellum. This spatial proteomics analysis provides a framework which can underpin current research and open new avenues to enhance molecular understanding of AD pathophysiology, provide new targets for intervention and broaden the conceptual frameworks for future AD research.

Despite all the significant advances in pedestrian detection brought by computer vision for driving assistance, it is still a challenging problem. One reason is the extremely varying lighting conditions under which such a detector should operate, namely day and nighttime. Recent research has shown that the combination of visible and non-visible imaging modalities may increase detection accuracy, where the infrared spectrum plays a critical role. The goal of this paper is to assess the accuracy gain of different pedestrian models (holistic, part-based, patch-based) when training with images in the far infrared spectrum. Specifically, we want to compare detection accuracy on test images recorded at day and nighttime if trained (and tested) using (a) plain color images; (b) just infrared images; and (c) both of them. In order to obtain results for the last item, we propose an early fusion approach to combine features from both modalities. We base the evaluation on a new dataset that we have built for this purpose as well as on the publicly available KAIST multispectral dataset.

Datasets comprising simultaneous measurements of many essential metals in Alzheimer's disease (AD) brain are sparse, and available studies are not entirely in agreement. To further elucidate this matter, we employed inductively-coupled-plasma mass spectrometry to measure post-mortem levels of 8 essential metals and selenium, in 7 brain regions from 9 cases with AD (neuropathological severity Braak IV-VI), and 13 controls who had normal ante-mortem mental function and no evidence of brain disease. Of the regions studied, three undergo severe neuronal damage in AD (hippocampus, entorhinal cortex and middle-temporal gyrus); three are less-severely affected (sensory cortex, motor cortex and cingulate gyrus); and one (cerebellum) is relatively spared. Metal concentrations in the controls differed among brain regions, and AD-associated perturbations in most metals occurred in only a few: regions more severely affected by neurodegeneration generally showed alterations in more metals, and cerebellum displayed a distinctive pattern. By contrast, copper levels were substantively decreased in all AD-brain regions, to 52.8-70.2% of corresponding control values, consistent with pan-cerebral copper deficiency. This copper deficiency could be pathogenic in AD, since levels are lowered to values approximating those in Menkes' disease, an X-linked recessive disorder where brain-copper deficiency is the accepted cause of severe brain damage. Our study reinforces others reporting deficient brain copper in AD, and indicates that interventions aimed at safely and effectively elevating brain copper could provide a new experimental-therapeutic approach.

Recent progress of self-supervised visual representation learning has achieved remarkable success on many challenging computer vision benchmarks. However, whether these techniques can be used for domain adaptation has not been explored. In this work, we propose a generic method for self-supervised domain adaptation, using object recognition and semantic segmentation of urban scenes as use cases. Focusing on simple pretext/auxiliary tasks (e.g. image rotation prediction), we assess different learning strategies to improve domain adaptation effectiveness by self-supervision. Additionally, we propose two complementary strategies to further boost the domain adaptation accuracy within our method, consisting of prediction layer alignment and batch normalization calibration. For the experimental work, we focus on the relevant setting of training models using synthetic images, and adapting them to perform on real-world images. The obtained results show adaptation levels comparable to most studied domain adaptation methods, thus, bringing self-supervision as a new alternative for reaching domain adaptation. The code is available at https://github.com/Jiaolong/self-supervised-da.

BackgroundHeart disease is the leading cause of death in diabetic patients, and defective copper metabolism may play important roles in the pathogenesis of diabetic cardiomyopathy (DCM). The present study sought to determine how myocardial copper status and key copper-proteins might become impaired by diabetes, and how they respond to treatment with the Cu (II)-selective chelator triethylenetetramine (TETA) in DCM.MethodsExperiments were performed in Wistar rats with streptozotocin (STZ)-induced diabetes with or without TETA treatment. Cardiac function was analyzed in isolated-perfused working hearts, and myocardial total copper content measured by particle-induced x-ray emission spectroscopy (PIXE) coupled with Rutherford backscattering spectrometry (RBS). Quantitative expression (mRNA and protein) and/or activity of key proteins that mediate LV-tissue-copper binding and transport, were analyzed by combined RT-qPCR, western blotting, immunofluorescence microscopy, and enzyme activity assays. Statistical analysis was performed using Student’s t-tests or ANOVA and p-values of < 0.05 have been considered significant.ResultsLeft-ventricular (LV) copper levels and function were severely depressed in rats following 16-weeks’ diabetes, but both were unexpectedly normalized 8-weeks after treatment with TETA was instituted. Localized myocardial copper deficiency was accompanied by decreased expression and increased polymerization of the copper-responsive transition-metal-binding metallothionein proteins (MT1/MT2), consistent with impaired anti-oxidant defences and elevated susceptibility to pro-oxidant stress. Levels of the high-affinity copper transporter-1 (CTR1) were depressed in diabetes, consistent with impaired membrane copper uptake, and were not modified by TETA which, contrastingly, renormalized myocardial copper and increased levels and cell-membrane localization of the low-affinity copper transporter-2 (CTR2). Diabetes also lowered indexes of intracellular (IC) copper delivery via the copper chaperone for superoxide dismutase (CCS) to its target cuproenzyme, superoxide dismutase-1 (SOD1): this pathway was rectified by TETA treatment, which normalized SOD1 activity with consequent bolstering of anti-oxidant defenses. Furthermore, diabetes depressed levels of additional intracellular copper-transporting proteins, including antioxidant-protein-1 (ATOX1) and copper-transporting-ATPase-2 (ATP7B), whereas TETA elevated copper-transporting-ATPase-1 (ATP7A).ConclusionsMyocardial copper deficiency and defective cellular copper transport/trafficking are revealed as key molecular defects underlying LV impairment in diabetes, and TETA-mediated restoration of copper regulation provides a potential new class of therapeutic molecules for DCM.

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that currently affects 36 million people worldwide with no effective treatment available. Development of AD follows a distinctive pattern in the brain and is poorly modelled in animals. Therefore, it is vital to widen both the spatial scope of the study of AD and prioritise the study of human brains. Here we show that functionally distinct human brain regions show varying and region-specific changes in protein expression. These changes provide novel insights into the progression of disease, novel AD-related pathways, the presence of a ‘gradient’ of protein expression change from less to more affected regions, and the presence of a ‘protective’ protein expression profile in the cerebellum. This spatial proteomics analysis provides a framework which can underpin current research and opens new avenues of interest to enhance our understanding of molecular pathophysiology of AD, provides new targets for intervention and broadens the conceptual frameworks for future AD research.

Abstract. We proposed a deep learning method for forest fire detection. We train both a full image and fine grained patch fire classifier in a joined deep convolutional neural networks (CNN). The fire detection is operated in a cascaded fashion, ie the full image is first tested by the global image-level classifier, if fire is detected, the fine grained patch classifier is followed to detect the precise location of fire patches. Our fire patch detector obtains 97% and 90% detection accuracy on training and testing datasets respectively. To facilitate the evaluation of various fire detectors in the community, we build a fire detection benchmark. According to our best knowledge, this is the first one with patch-level annotations.