US-Cut: Interactive Algorithm for rapid Detection and Segmentation of Liver Tumors in Ultrasound Acquisitions

摘要

Ultrasound (US) is the most commonly used liver imaging modality worldwide.It plays an important role in follow-up of cancer patients with livermetastases. We present an interactive segmentation approach for liver tumors inUS acquisitions. Due to the low image quality and the low contrast between thetumors and the surrounding tissue in US images, the segmentation is verychallenging. Thus, the clinical practice still relies on manual measurement andoutlining of the tumors in the US images. We target this problem by applying aninteractive segmentation algorithm to the US data, allowing the user to getreal-time feedback of the segmentation results. The algorithm has beendeveloped and tested hand-in-hand by physicians and computer scientists to makesure a future practical usage in a clinical setting is feasible. To covertypical acquisitions from the clinical routine, the approach has been evaluatedwith dozens of datasets where the tumors are hyperechoic (brighter), hypoechoic(darker) or isoechoic (similar) in comparison to the surrounding liver tissue.Due to the interactive real-time behavior of the approach, it was possible evenin difficult cases to find satisfying segmentations of the tumors withinseconds and without parameter settings, and the average tumor deviation wasonly 1.4mm compared with manual measurements. However, the long term goal is toease the volumetric acquisition of liver tumors in order to evaluate fortreatment response. Additional aim is the registration of intraoperative USimages via the interactive segmentations to the patient's pre-interventional CTacquisitions.