Artificial intelligence (AI) has shown promise for diagnosing prostate cancer in biopsies. However, results have been limited to individual studies, lacking validation in multinational settings. Competitions have been shown to be accelerators for medical imaging innovations, but their impact is hindered by lack of reproducibility and independent validation. With this in mind, we organized the PANDA challenge—the largest histopathology competition to date, joined by 1,290 developers—to catalyze development of reproducible AI algorithms for Gleason grading using 10,616 digitized prostate biopsies. We validated that a diverse set of submitted algorithms reached pathologist-level performance on independent cross-continental cohorts, fully blinded to the algorithm developers. On United States and European external validation sets, the algorithms achieved agreements of 0.862 (quadratically weighted κ, 95% confidence interval (CI), 0.840–0.884) and 0.868 (95% CI, 0.835–0.900) with expert uropathologists. Successful generalization across different patient populations, laboratories and reference standards, achieved by a variety of algorithmic approaches, warrants evaluating AI-based Gleason grading in prospective clinical trials.

The Gleason score is the most important prognostic marker for prostate cancer patients, but it suffers from significant observer variability. Artificial intelligence (AI) systems based on deep learning can achieve pathologist-level performance at Gleason grading. However, the performance of such systems can degrade in the presence of artifacts, foreign tissue, or other anomalies. Pathologists integrating their expertise with feedback from an AI system could result in a synergy that outperforms both the individual pathologist and the system. Despite the hype around AI assistance, existing literature on this topic within the pathology domain is limited. We investigated the value of AI assistance for grading prostate biopsies. A panel of 14 observers graded 160 biopsies with and without AI assistance. Using AI, the agreement of the panel with an expert reference standard increased significantly (quadratically weighted Cohen's kappa, 0.799 vs. 0.872; p = 0.019). On an external validation set of 87 cases, the panel showed a significant increase in agreement with a panel of international experts in prostate pathology (quadratically weighted Cohen's kappa, 0.733 vs. 0.786; p = 0.003). In both experiments, on a grouplevel, AI-assisted pathologists outperformed the unassisted pathologists and the standalone AI system. Our results show the potential of AI systems for Gleason grading, but more importantly, show the benefits of pathologist-AI synergy. Members of the ISUP Pathology Imagebase Expert Panel are listed below Acknowledgements.

Abstract. São Paulo in Brazil has relatively relaxed regulations for ambient air pollution standards and often experiences high air pollution levels due to emissions of particulate pollutants from local sources and long-range transport of air masses impacted by biomass burning. In order to evaluate the sources of particulate air pollution and related health risks, a year-round sampling was done at the University of São Paulo campus (20 m a.g.l.), a green area near an important expressway. The sampling was performed for PM 2.5 (≤ 2.5 µm) and PM 10 (≤ 10 µm) in 2014 through intensive (everyday sampling in wintertime) and extensive campaigns (once a week for the whole year) with 24 h of sampling. This year was characterized by having lower average precipitation compared to meteorological data, and high-pollution episodes were observed all year round, with a significant increase in pollution level in the intensive campaign, which was performed during wintertime. Different chemical constituents, such as carbonaceous species, polycyclic aromatic hydrocarbons (PAHs) and derivatives, water-soluble ions, and biomass burning tracers were identified in order to evaluate health risks and to apportion sources. The species such as PAHs, inorganic and organic ions, and monosaccharides were determined using chromatographic techniques and carbonaceous species using thermal-optical analysis. Trace elements were determined using inductively coupled plasma mass spectrometry. The risks associated with particulate matter exposure based on PAH concentrations were also assessed, along with indexes such as the benzo[a]pyrene equivalent (BaPE) and lung cancer risk (LCR). High BaPE and LCR were observed in most of the samples, rising to critical values in the wintertime. Also, biomass burning tracers and PAHs were higher in this season, while secondarily formed ions presented low variation throughout the year. Meanwhile, vehicular tracer species were also higher in the intensive campaign, suggesting the influence of lower dispersion conditions in that period. Source apportionment was performed using positive matrix factorization (PMF), which indicated five different factors: road dust, industrial emissions, vehicuPublished by Copernicus Publications on behalf of the European Geosciences Union. 11944 G. M. Pereira et al.: Particulate pollutants in the Brazilian city of São Paulo lar exhaust, biomass burning and secondary processes. The results highlighted the contribution of vehicular emissions and the significant input from biomass combustion in wintertime, suggesting that most of the particulate matter is due to local sources, in addition to the influence of pre-harvest sugarcane burning.

No abstract

No abstract