3D Statistical Shape Models Incorporating Landmark-Wise Random Regression Forests for Omni-Directional Landmark Detection

Abstract

3D Statistical Shape Models (3D-SSM) are widely used for medical image segmentation. However, during segmentation, they typically perform a very limited unidirectional search for suitable landmark positions in the image, relying on weak learners or use-case specific appearance models that solely take local image information into account. As a consequence, segmentation errors arise, and results in general depend on the accuracy of a previous model initialization. Furthermore, these methods become subject to a tedious and use-case dependent parameter tuning in order to obtain optimized results. To overcome these limitations, we propose an extension of 3D-SSM by landmark-wise random regression forests that perform an enhanced omni-directional search for landmark positions, thereby taking rich non-local image information into account. In addition, we provide a long distance model fitting based on a multi-scale approach, that allows an accurate and reproducible segmentation even from distant image positions, thus enabling an application without model initialization. Finally, translation of the proposed method to different organs is straightforward and requires no adaptation of the training process. In segmentation experiments on 45 clinical CT volumes, the proposed omni-directional search significantly increased accuracy and displayed great precision regardless of model initialization. Furthermore, for liver, spleen and kidney segmentation in a competitive multi-organ labeling challenge on publicly available data, the proposed method achieved similar or better results than the state of the art. Finally, liver segmentation results were obtained that successfully compete with specialized state-of-the-art methods from the well-known liver segmentation challenge SLIVER.

Soft-tissue-segmentation methods during image-guided precision liver surgery

“…Although the segmentation methods based on the deformation model, statistical shape model and probabilistic atlas model have been widely used in the field of liver segmentation and tested to be effective, the following problems remain to be addressed [ 3 ]. (i) In the deformation model, the validity and robustness of the internal and external force constraint models must be improved.…”

Intravoxel Incoherent Motion

“…In addition, it must be said that, over the past years, many different computational techniques have been described for organ segmentation. 26 This may raise the question, "Why use IVIM data sets for tissue type classification and segmentation?" Because most segmentation techniques are based on surface and landmark detection and typically not relying on organ characteristics except for their shape, these algorithms may run into serious problems for classification, if the anatomy of the organs is severely changed.…”

Strategies for prediction and mitigation of radiation-induced liver toxicity

“…Machine learning–based methods have been developed for automated analysis of the majority of OARs that may be involved in liver SBRT. Because of the availability of a public database and unified evaluation software, liver segmentation has been receiving considerable attention in the literature [ 100 , 101 , 103 – 110 ]. The existing machine learning–based methods exhibit liver segmentation accuracy of from 67 to 97.3%, measured in terms of Dice coefficient, and ~5% of volume disagreement from that obtained with manual segmentation [ 107 , 108 ].…”