Advancements in volume electron microscopy mean it is now possible to generate thousands of serial images at nanometre resolution overnight, yet the gold standard approach for data analysis remains manual segmentation by an expert microscopist, resulting in a critical research bottleneck. Although some machine learning approaches exist in this domain, we remain far from realizing the aspiration of a highly accurate, yet generic, automated analysis approach, with a major obstacle being lack of sufficient high‐quality ground‐truth data. To address this, we developed a novel citizen science project, Etch a Cell, to enable volunteers to manually segment the nuclear envelope (NE) of HeLa cells imaged with serial blockface scanning electron microscopy. We present our approach for aggregating multiple volunteer annotations to generate a high‐quality consensus segmentation and demonstrate that data produced exclusively by volunteers can be used to train a highly accurate machine learning algorithm for automatic segmentation of the NE, which we share here, in addition to our archived benchmark data.

Advancements in volume electron microscopy mean it is now possible to generate thousands of serial images at nanometre resolution overnight, yet the gold standard approach for data analysis remains manual segmentation by an expert microscopist, resulting in a critical research bottleneck. Although some machine learning approaches exist in this domain, we remain far from realising the aspiration of a highly accurate, yet generic, automated analysis approach, with a major obstacle being lack of sufficient high-quality ground-truth data. To address this, we developed a novel citizen science project, Etch a Cell, to enable volunteers to manually segment the nuclear envelope of HeLa cells imaged with Serial Blockface SEM. We present our approach for aggregating multiple volunteer annotations to generate a high quality consensus segmentation, and demonstrate that data produced exclusively by volunteers can be used to train a highly accurate machine learning algorithm for automatic segmentation of the nuclear envelope, which we share here, in addition to our archived benchmark data.

The ongoing pandemic of SARS-CoV-2 calls for rapid and cost-effective methods to accurately identify infected individuals. The vast majority of patient samples is assessed for viral RNA presence by RT-qPCR. Our biomedical research institute, in collaboration between partner hospitals and an accredited clinical diagnostic laboratory, established a diagnostic testing pipeline that has reported on more than 40,000 RT-qPCR results since its commencement at the beginning of April 2020. However, due to ongoing demand and competition for critical resources, alternative testing strategies were sought. In this work, we present a clinically-validated standard operating procedure (SOP) for high-throughput SARS- CoV-2 detection by RT-LAMP in 25 minutes that is robust, reliable, repeatable, sensitive, specific, and inexpensive.

The emergence of the novel coronavirus SARS-CoV-2 has led to a pandemic infecting more than two million people worldwide in less than four months, posing a major threat to healthcare systems. This is compounded by the shortage of available tests causing numerous healthcare workers to unnecessarily self-isolate. We provide a roadmap instructing how a research institute can be repurposed in the midst of this crisis, in collaboration with partner hospitals and an established diagnostic laboratory, harnessing existing expertise in virus handling, robotics, PCR, and data science to derive a rapid, high throughput diagnostic testing pipeline for detecting SARS-CoV-2 in patients with suspected COVID-19. The pipeline is used to detect SARS-CoV-2 from combined nose-throat swabs and endotracheal secretions/ bronchoalveolar lavage fluid. Notably, it relies on a series of in-house buffers for virus inactivation and the extraction of viral RNA, thereby reducing the dependency on commercial suppliers at times of global shortage. We use a commercial RT-PCR assay, from BGI, and results are reported with a bespoke online web application that integrates with the healthcare digital system. This strategy facilitates the remote reporting of thousands of samples a day with a turnaround time of under 24 hours, universally applicable to laboratories worldwide.

The ongoing pandemic of SARS-CoV-2 calls for rapid and cost-effective methods to accurately identify infected individuals. The vast majority of patient samples is assessed for viral RNA presence by RT-qPCR. Our biomedical research institute, in collaboration between partner hospitals and an accredited clinical diagnostic laboratory, established a diagnostic testing pipeline that has reported on more than 252,000 RT-qPCR results since its commencement at the beginning of April 2020. However, due to ongoing demand and competition for critical resources, alternative testing strategies were sought. In this work, we present a clinically-validated procedure for high-throughput SARS-CoV-2 detection by RT-LAMP that is robust, reliable, repeatable, specific, and inexpensive.