Background: Deep learning enables accurate high-resolution mapping of cells and tissue structures that can serve as the foundation of interpretable machine-learning models for computational pathology. However, generating adequate labels for these structures is a critical barrier, given the time and effort required from pathologists. Results: This article describes a novel collaborative framework for engaging crowds of medical students and pathologists to produce quality labels for cell nuclei. We used this approach to produce the NuCLS dataset, containing >220,000 annotations of cell nuclei in breast cancers. This builds on prior work labeling tissue regions to produce an integrated tissue region- and cell-level annotation dataset for training that is the largest such resource for multi-scale analysis of breast cancer histology. This article presents data and analysis results for single and multi-rater annotations from both non-experts and pathologists. We present a novel workflow that uses algorithmic suggestions to collect accurate segmentation data without the need for laborious manual tracing of nuclei. Our results indicate that even noisy algorithmic suggestions do not adversely affect pathologist accuracy and can help non-experts improve annotation quality. We also present a new approach for inferring truth from multiple raters and show that non-experts can produce accurate annotations for visually distinctive classes. Conclusions: This study is the most extensive systematic exploration of the large-scale use of wisdom-of-the-crowd approaches to generate data for computational pathology applications.
6 September 2019 in International Journal of Colorectal Disease
Background: Postpartum Hemorrhage (PPH) is one of the primary causes of maternal mortality and morbidity during the third stage of labor. Oxytocin is the gold standard uterotonic agent for the prevention of PPH. Objective: We aimed to compare the efficacy of oxytocin administered Intramuscularly (IM) or Intravenously (IV) for the preventive management of PPH. Methods: We searched six databases for relevant clinical trials evaluating the administration of oxytocin for the prevention against PPH through July 2019. Data on blood loss, PPH (≥500 ml), severe PPH (≥1000 ml), blood transfusion, the change in hemoglobin, the use of additional uterotonics, and the incidence of retained placenta were extracted and pooled in a meta-analysis model using RevMan version 5.3. Results: Seven studies with a total of 6996 participants were included. IM oxytocin group was associated with higher incidence rates of PPH (≥500 ml) (RR=1.35; p=0.003), severe PPH (≥1000 ml) (RR=1.58; p=0.04), and blood transfusion (RR=2.43; p=0.005). In terms of blood loss, the IV route was superior to the IM route (SMD= 0.15; p=0.00001). However, we observed no statistically significant difference between the two routes regarding the change in Hb (SMD=-0.02; p=0.72) and the use of additional uterotonics (RR=0.96, p= 0.94). Conclusion: IV oxytocin infusion is maybe superior to IM injection for the management of PPH. Further studies with larger sample sizes are still needed to support these findings.