VoteNet+ : An Improved Deep Learning Label Fusion Method for Multi-atlas Segmentation
This is an incremental improvement for medical image segmentation, specifically in brain MRI analysis.
The authors tackled the problem of multi-atlas segmentation by integrating VoteNet with joint label fusion, resulting in improved performance on LPBA40 3D MR brain images compared to VoteNet.
In this work, we improve the performance of multi-atlas segmentation (MAS) by integrating the recently proposed VoteNet model with the joint label fusion (JLF) approach. Specifically, we first illustrate that using a deep convolutional neural network to predict atlas probabilities can better distinguish correct atlas labels from incorrect ones than relying on image intensity difference as is typical in JLF. Motivated by this finding, we propose VoteNet+, an improved deep network to locally predict the probability of an atlas label to differs from the label of the target image. Furthermore, we show that JLF is more suitable for the VoteNet framework as a label fusion method than plurality voting. Lastly, we use Platt scaling to calibrate the probabilities of our new model. Results on LPBA40 3D MR brain images show that our proposed method can achieve better performance than VoteNet.