A Practical Roadmap to Implementing Deep Learning Segmentation in the Clinical Neuroimaging Research Workflow.

BACKGROUND: Thanks to the proliferation of open-source tools, we are seeing an exponential growth of machine-learning applications, and its integration has become more accessible, particularly for segmentation tools in neuroimaging. METHODS: This article explores a generalized methodology that harnesses these tools and aims/enables to expedite and enhance the reproducibility of clinical research. Herein, critical considerations include hardware, software, neural network training strategies, and data labeling guidelines. More specifically, we advocate an iterative approach to model training and transfer learning, focusing on internal validation and outlier handling early in the labeling process and fine-tuning later. RESULTS: The iterative refinement process allows experts to intervene and improve model reliability while cutting down on their time spent in manual work. A seamless integration of the final model's predictions into clinical research is proposed to ensure standardized and reproducible results. CONCLUSIONS: In short, this article provides a comprehensive framework for accelerating research using machine-learning techniques for image segmentation.

Fate and Trophic Transfer of Rare Earth Elements in Temperate Lake Food Webs.

Many mining projects targeting rare earth elements (REE) are in development in North America, but the background concentrations and trophic transfer of these elements in natural environments have not been well characterized. We sampled abiotic and food web components in 14 Canadian temperate lakes unaffected by mines to assess the natural ecosystem fate of REE. Individual REE and total REE concentrations (sum of individual element concentrations, ΣREE) were strongly related with each other throughout different components of lake food webs. Dissolved organic carbon and dissolved oxygen in the water column, as well as ΣREE in sediments, were identified as potential drivers of aqueous ΣREE. Log10 of median bioaccumulation factors ranged from 1.3, 3.7, 4.0, and 4.4 L/kg (wet weight) for fish muscle, zooplankton, predatory invertebrates, and nonpredatory invertebrates, respectively. [ΣREE] in fish, benthic macroinvertebrates, and zooplankton declined as a function of their trophic position, as determined by functional feeding groups and isotopic signatures of nitrogen (δ15N), indicating that REE were subject to trophic dilution. Low concentrations of REE in freshwater fish muscle compared to their potential invertebrate prey suggest that fish fillet consumption is unlikely to be a significant source of REE to humans in areas unperturbed by mining activities. However, other fish predators (e.g., piscivorous birds and mammals) may accumulate REE from whole fish as they are more concentrated than muscle. Overall, this study provides key information on the baseline concentrations and trophic patterns for REE in freshwater temperate lakes in Quebec, Canada.