Spatial bias in multi-atlas based segmentation

Multi-atlas segmentation has been widely applied in medical image analysis. With deformable registration, this technique realizes label transfer from pre-labeled atlases to unknown images. When deformable registration produces error, label fusion that combines results produced by multiple atlases is an effective way for reducing segmentation errors. Among the existing label fusion strategies, similarityweighted voting strategies with spatially varying weight distributions have been particularly successful. We show that, weighted voting based label fusion produces a spatial bias that under-segments structures with convex shapes. The bias can be approximated as applying spatial convolution to the ground truth spatial label probability maps, where the convolution kernel combines the distribution of residual registration errors and the function producing similaritybased voting weights. To reduce this bias, we apply a standard spatial deconvolution to the spatial probability maps obtained ...

• Type: Journal
• Year: 2012
• Where: CVPR
• Authors: Hongzhi Wang, Paul A. Yushkevich
• Keywords: Computer Vision, Convolution Kernel, Cvpr 2012, Medical Image Analysis, Probability Maps