Abstract: Models that produce axionlike particles (ALPs) after cosmological inflation due to spontaneous U(1) symmetry breaking also produce cosmic-string networks. Those axionic strings lose energy through gravitational-wave emission during the whole cosmological history, generating a stochastic background of gravitational waves that spans many decades in frequency. We can therefore constrain the axion decay constant and axion mass from limits on the gravitational-wave spectrum and compatibility with dark matter abundance as well as dark radiation. We derive such limits from analyzing the most recent NANOGrav data from pulsar timing arrays (PTAs). The limits are similar to the Neff bounds on dark radiation for ALP masses ma less than or similar to 10-22 eV. On the other hand, for heavy ALPs with ma greater than or similar to 0.1 GeV and NDW not equal 1, new regions of parameter space can be probed by PTA data due to the dominant domain-wall contribution to the gravitational-wave background.

Abstract: For the Belle II experiment at KEK (Tsukuba, Japan) the KEKB accelerator was upgraded to deliver a 40 times larger instantaneous luminosity than before, which requires an increased radiation hardness of the detector components. As the innermost part of the Belle II detector, the pixel detector (PXD), based on DEPFET (DEpleted P-channel Field Effect Transistor) technology, is most exposed to radiation from the accelerator. An irradiation campaign was performed to verify that the PXD can cope with the expected amount of radiation. We present the results of this measurement campaign in which an X-ray machine was used to irradiate a single PXD half-ladder to a total dose of 266 kGy. The half-ladder is from the same batch as the half-ladders used for Belle II. According to simulations, the total accumulated dose corresponds to 7-10 years of Belle II operation. While individual components have been irradiated before, this campaign is the first full system irradiation. We discuss the effects on the DEPFET sensors, as well as the performance of the front-end electronics. In addition, we present efficiency studies of the half-ladder from beam tests performed before and after the irradiation.

Abstract: Importance Mammography screening currently relies on subjective human interpretation. Artificial intelligence (AI) advances could be used to increase mammography screening accuracy by reducing missed cancers and false positives. Objective To evaluate whether AI can overcome human mammography interpretation limitations with a rigorous, unbiased evaluation of machine learning algorithms. Design, Setting, and Participants In this diagnostic accuracy study conducted between September 2016 and November 2017, an international, crowdsourced challenge was hosted to foster AI algorithm development focused on interpreting screening mammography. More than 1100 participants comprising 126 teams from 44 countries participated. Analysis began November 18, 2016. Main Outcomes and Measurements Algorithms used images alone (challenge 1) or combined images, previous examinations (if available), and clinical and demographic risk factor data (challenge 2) and output a score that translated to cancer yes/no within 12 months. Algorithm accuracy for breast cancer detection was evaluated using area under the curve and algorithm specificity compared with radiologists' specificity with radiologists' sensitivity set at 85.9% (United States) and 83.9% (Sweden). An ensemble method aggregating top-performing AI algorithms and radiologists' recall assessment was developed and evaluated. Results Overall, 144231 screening mammograms from 85580 US women (952 cancer positive <= 12 months from screening) were used for algorithm training and validation. A second independent validation cohort included 166578 examinations from 68008 Swedish women (780 cancer positive). The top-performing algorithm achieved an area under the curve of 0.858 (United States) and 0.903 (Sweden) and 66.2% (United States) and 81.2% (Sweden) specificity at the radiologists' sensitivity, lower than community-practice radiologists' specificity of 90.5% (United States) and 98.5% (Sweden). Combining top-performing algorithms and US radiologist assessments resulted in a higher area under the curve of 0.942 and achieved a significantly improved specificity (92.0%) at the same sensitivity. Conclusions and Relevance While no single AI algorithm outperformed radiologists, an ensemble of AI algorithms combined with radiologist assessment in a single-reader screening environment improved overall accuracy. This study underscores the potential of using machine learning methods for enhancing mammography screening interpretation. Question How do deep learning algorithms perform compared with radiologists in screening mammography interpretation? Findings In this diagnostic accuracy study using 144231 screening mammograms from 85580 women from the United States and 166578 screening mammograms from 68008 women from Sweden, no single artificial intelligence algorithm outperformed US community radiologist benchmarks; including clinical data and prior mammograms did not improve artificial intelligence performance. However, combining best-performing artificial intelligence algorithms with single-radiologist assessment demonstrated increased specificity. Meaning Integrating artificial intelligence to mammography interpretation in single-radiologist settings could yield significant performance improvements, with the potential to reduce health care system expenditures and address resource scarcity experienced in population-based screening programs. This diagnostic accuracy study evaluates whether artificial intelligence can overcome human mammography interpretation limits with a rigorous, unbiased evaluation of machine learning algorithms.