Segmentation of 4D images via space-time neural networks

摘要

Medical imaging techniques currently produce 4D images that portray the dynamic behaviors and phenomena associatedwith internal structures. The segmentation of 4D images poses challenges different from those arising in segmenting 3Dstatic images due to different patterns of variation of object shape and appearance in the space and time dimensions. Inthis paper, different network models are designed to learn the pattern of slice-to-slice change in the space and timedimensions independently. The two models then allow a gamut of strategies to actually segment the 4D image, such assegmentation following just the space or time dimension only, or following first the space dimension for one time instanceand then following all time instances, or vice versa, etc. This paper investigates these strategies in the context of theobstructive sleep apnea (OSA) application and presents a unified deep learning framework to segment 4D images. Becauseof the sparse tubular nature of the upper airway and the surrounding low-contrast structures, inadequate contrast resolutionobtainable in the magnetic resonance (MR) images leaves many challenges for effective segmentation of the dynamicairway in 4D MR images. Given that these upper airway structures are sparse, a Dice coefficient (DC) of ～0.88 for theirsegmentation based on our preferred strategy is similar to a DC of >0.95 for large non-sparse objects like liver, lungs, etc.,constituting excellent accuracy.