Active Scheduling of Organ Detection and Segmentation in Whole-Body Medical Images

With the advance of whole-body medical imaging technologies, computer aided detection/diagnosis (CAD) is being scaled up to deal with multiple organs or anatomical structures simultaneously. Multiple tasks (organ detection/segmentation) in a CAD system are often highly dependent due to the anatomical context within a human body. In this paper, we propose a method to schedule multi-organ detection/segmentation based on information theory. The central idea is to schedule tasks in an order that each operation achieves maximum expected information gain. The scheduling rule is formulated to embed two intuitive principles: (1) a task with higher confidence tends to be scheduled earlier; (2) a task with higher predictive power for other tasks tends to be scheduled earlier. More specifically, task dependency is modeled by conditional probability; the outcome of each task is assumed to be probabilistic as well; and the scheduling criterion is based on the reduction of the summed conditional entropy over all tasks. The validation is carried out on two challenging CAD problems, multi-organ detection in whole-body CT and liver segmentation in PET-CT. Compared to unscheduled and ad hoc scheduled organ detection/segmentation, our scheduled execution achieves higher accuracy with faster speed.