Benchmarking of deep learning algorithms for 3D instance segmentation of confocal image datasets

Abstract

Segmenting three dimensional microscopy images is essential for understanding phenomena like morphogenesis, cell division, cellular growth and genetic expression patterns. Recently, deep learning (DL) pipelines have been developed which claim to provide high accuracy segmentation of cellular images and are increasingly considered as the state-of-the-art for image segmentation problems. However, it remains difficult to define their relative performance as the concurrent diversity and lack of uniform evaluation strategies makes it difficult to know how their results compare. In this paper, we first made an inventory of the available DL methods for 3D segmentation. We next implemented and quantitatively compared a number of representative DL pipelines, alongside a highly efficient non-DL method named MARS. The DL methods were trained on a common dataset of 3D cellular confocal microscopy images. Their segmentation accuracies were also tested in the presence of different image artefacts. A new method for segmentation quality evaluation was adopted which isolates segmentation errors due to under/over segmentation. This is complemented with new visualisation strategies that make interactive exploration of segmentation quality possible. Our analysis shows that the DL pipelines have very different levels of accuracy. Two of them show high performance, and offer clear advantages in terms of adaptability to new data.