interactive 3D object segmentation is an important and challenging activity in medical imaging, although it is tedious and error-prone to be done. Automatic segmentation methods aim to replace the user altogether, but...
详细信息
ISBN:
(纸本)9781509035687
interactive 3D object segmentation is an important and challenging activity in medical imaging, although it is tedious and error-prone to be done. Automatic segmentation methods aim to replace the user altogether, but require user interaction to produce training data sets of segmented masks and to make error corrections. We propose a complete framework for interactive medical image segmentation, which reduces user effort by automatically providing an initial segmentation result. We develop a Statistical Seed Model (SSM) to this end, that improves from seed sets selected by robot users when reconstructing masks of previously segmented images. The SSM outputs a seed set that may be used to automatically delineate a new test image. The seeds provide both an implicit object shape constraint and a flexible way of interactively correcting segmentation. We demonstrate that our framework decreases the amount of user interaction by a factor of three, when segmenting MR-images of the cerebellum.
The segmentation Anything Model (SAM) has recently emerged as a foundation model for addressing imagesegmentation. Owing to the intrinsic complexity of medicalimages and the high annotation cost, the medicalimage s...
详细信息
Deep learning-based interactivesegmentation has attracted research interest recently since it can smartly utilize user interactions to refine a coarse automatic segmentation to get higher accuracy for clinical use. C...
详细信息
ISBN:
(数字)9783030598617
ISBN:
(纸本)9783030598600;9783030598617
Deep learning-based interactivesegmentation has attracted research interest recently since it can smartly utilize user interactions to refine a coarse automatic segmentation to get higher accuracy for clinical use. Current methods usually transform user clicks to geodesic distance hint maps as guidance, then concatenate them with the raw image and coarse segmentation, and feed them into a refinement network. Such methods are insufficient in refining error region, which is a key capability required for interactivesegmentation. In this paper, we propose Error Attention interactive network with Matting and Fusion to autoextract guide information of mis-segmentation region from two branches and transfer it into main segmentor. We first design Region Matting to obtain foreground and background mattings from coarse segmentation. And then we adopt the features extracted by two branches trained on above mattings as guidance. Attention-Fusion is further proposed to transfer the guidance to main segmentor effectively based on attention mechanism and feature concatenation. Experimental results on BraTS 2015 and our Neuroblastoma datasets have shown that our method significantly outperforms state-of-the-art methods, with the advantage of fewer interactions.
This paper describes a new method for interactivesegmentation that is based on cross-sectional design and 3D modelling. The method represents a 3D model by a set of connected contours that are planar and orthogonal. ...
详细信息
This paper describes a new method for interactivesegmentation that is based on cross-sectional design and 3D modelling. The method represents a 3D model by a set of connected contours that are planar and orthogonal. Planar contours overlayed on image data are easily manipulated and linked contours reduce the amount of user interaction. This method solves the contour-to-contour correspondence problem and can capture extrema of objects in a more flexible way than manual segmentation of a stack of 2D images. The resulting 3D model is guaranteed to be free of geometric and topological errors. We show that manual segmentation using connected orthogonal contours has great advantages over conventional manual segmentation. Furthermore, the method provides effective feedback and control for creating an initial model for, and control and steering of, (semi-)automatic segmentation methods. (c) 2004 Elsevier Ltd. All rights reserved.
暂无评论