资源环境科技发展态势分析平台

Screening mammography aims to identify breast cancer at earlier stages of the disease, when treatment can be more successful(1). Despite the existence of screening programmes worldwide, the interpretation of mammograms is affected by high rates of false positives and false negatives(2). Here we present an artificial intelligence (AI) system that is capable of surpassing human experts in breast cancer prediction. To assess its performance in the clinical setting, we curated a large representative dataset from the UK and a large enriched dataset from the USA. We show an absolute reduction of 5.7% and 1.2% (USA and UK) in false positives and 9.4% and 2.7% in false negatives. We provide evidence of the ability of the system to generalize from the UK to the USA. In an independent study of six radiologists, the AI system outperformed all of the human readers: the area under the receiver operating characteristic curve (AUC-ROC) for the AI system was greater than the AUC-ROC for the average radiologist by an absolute margin of 11.5%. We ran a simulation in which the AI system participated in the double-reading process that is used in the UK, and found that the AI system maintained non-inferior performance and reduced the workload of the second reader by 88%. This robust assessment of the AI system paves the way for clinical trials to improve the accuracy and efficiency of breast cancer screening.

Lidar survey of the Maya lowlands uncovers the monumental site of Aguada Fenix, which dates to around 1000-800 bc and points to the role of communal construction in the development of Maya civilization.

An efficient entanglement-based quantum key distribution is sent from the Micius satellite to two ground observatories 1,120 kilometres apart to establish secure quantum cryptography for the exchange ofquantum keys.

Quantum key distribution (QKD)(1-3)is a theoretically secure way of sharing secret keys between remote users. It has been demonstrated in a laboratory over a coiled optical fibre up to 404 kilometres long(4-7). In the field, point-to-point QKD has been achieved from a satellite to a ground station up to 1,200 kilometres away(8-10). However, real-world QKD-based cryptography targets physically separated users on the Earth, for which the maximum distance has been about 100 kilometres(11,12). The use of trusted relays can extend these distances from across a typical metropolitan area(13-16)to intercity(17)and even intercontinental distances(18). However, relays pose security risks, which can be avoided by using entanglement-based QKD, which has inherent source-independent security(19,20). Long-distance entanglement distribution can be realized using quantum repeaters(21), but the related technology is still immature for practical implementations(22). The obvious alternative for extending the range of quantum communication without compromising its security is satellite-based QKD, but so far satellite-based entanglement distribution has not been efficient(23)enough to support QKD. Here we demonstrate entanglement-based QKD between two ground stations separated by 1,120 kilometres at a finite secret-key rate of 0.12 bits per second, without the need for trusted relays. Entangled photon pairs were distributed via two bidirectional downlinks from the Micius satellite to two ground observatories in Delingha and Nanshan in China. The development of a high-efficiency telescope and follow-up optics crucially improved the link efficiency. The generated keys are secure for realistic devices, because our ground receivers were carefully designed to guarantee fair sampling and immunity to all known side channels(24,25). Our method not only increases the secure distance on the ground tenfold but also increases the practical security of QKD to an unprecedented level.

Animals sense the environment through pathways that link sensory organs to the brain. In the visual system, these feedforward pathways define the classical feedforward receptive field (ffRF), the area in space in which visual stimuli excite a neuron(1). The visual system also uses visual context-the visual scene surrounding a stimulus-to predict the content of the stimulus(2), and accordingly, neurons have been identified that are excited by stimuli outside their ffRF(3-8). However, the mechanisms that generate excitation to stimuli outside the ffRF are unclear. Here we show that feedback projections onto excitatory neurons in the mouse primary visual cortex generate a second receptive field that is driven by stimuli outside the ffRF. The stimulation of this feedback receptive field (fbRF) elicits responses that are slower and are delayed in comparison with those resulting from the stimulation of the ffRF. These responses are preferentially reduced by anaesthesia and by silencing higher visual areas. Feedback inputs from higher visual areas have scattered receptive fields relative to their putative targets in the primary visual cortex, which enables the generation of the fbRF. Neurons with fbRFs are located in cortical layers that receive strong feedback projections and are absent in the main input layer, which is consistent with a laminar processing hierarchy. The observation that large, uniform stimuli-which cover both the fbRF and the ffRF-suppress these responses indicates that the fbRF and the ffRF are mutually antagonistic. Whereas somatostatin-expressing inhibitory neurons are driven by these large stimuli, inhibitory neurons that express parvalbumin and vasoactive intestinal peptide have mutually antagonistic fbRF and ffRF, similar to excitatory neurons. Feedback projections may therefore enable neurons to use context to estimate information that is missing from the ffRF and to report differences in stimulus features across visual space, regardless of whether excitation occurs inside or outside the ffRF. By complementing the ffRF, the fbRF that we identify here could contribute to predictive processing.

Feedback projections onto neurons of the mouse primary visual cortex generate a second excitatory receptive field that is driven by stimuli outside of the classical feedforward receptive field, with responses mediated by higher visual areas.

Since its introduction, the reward prediction error theory of dopamine has explained a wealth of empirical phenomena, providing a unifying framework for understanding the representation of reward and value in the brain(1-3). According to the now canonical theory, reward predictions are represented as a single scalar quantity, which supports learning about the expectation, or mean, of stochastic outcomes. Here we propose an account of dopamine-based reinforcement learning inspired by recent artificial intelligence research on distributional reinforcement learning(4-6). We hypothesized that the brain represents possible future rewards not as a single mean, but instead as a probability distribution, effectively representing multiple future outcomes simultaneously and in parallel. This idea implies a set of empirical predictions, which we tested using single-unit recordings from mouse ventral tegmental area. Our findings provide strong evidence for a neural realization of distributional reinforcement learning.

Analyses of single-cell recordings from mouse ventral tegmental area are consistent with a model of reinforcement learning in which the brain represents possible future rewards not as a single mean of stochastic outcomes, as in the canonical model, but instead as a probability distribution.