Large-scale imaging of locus coeruleus

Imaging the brainstem nucleus locus coeruleus (LC) in humans is of great interest for both scientific and clinical research. In recent years, multiple proposals for automatic segmentation of LC on magnetic resonance imaging sequences show potential for application to large samples. This article aims to review recent work segmenting the LC on MRI images from up to several hundreds of individuals. We identify relevant methodological factors that become particularly important for large samples as compared to smaller samples or specific studies. We survey the characteristics of the approaches, the reported reliability, and key methodological challenges. Overall, automatic segmentation algorithms can show comparable reliability to that obtained through manual segmentation, and show promise for their use in large datasets and the identification of candidate biomarkers of disease.