Feng Zishun, Nie Dong, Wang Li, Shen Dinggang
Department of Automation, Tsinghua University.
Department of Radiology and BRIC, UNC-Chapel Hill.
Proc IEEE Int Symp Biomed Imaging. 2018 Apr;2018:885-888. doi: 10.1109/ISBI.2018.8363713. Epub 2018 May 24.
Accurate segmentation of pelvic organs from magnetic resonance (MR) images plays an important role in image-guided radiotherapy. However, it is a challenging task due to inconsistent organ appearances and large shape variations. Fully convolutional network (FCN) has recently achieved state-of-the-art performance in medical image segmentation, but it requires a large amount of labeled data for training, which is usually difficult to obtain in real situation. To address these challenges, we propose a deep learning based semi-supervised learning framework. Specifically, we first train an initial multi-task residual fully convolutional network (FCN) based on a limited number of labeled MRI data. Based on the initially trained FCN, those unlabeled new data can be automatically segmented and some reasonable segmentations (after manual/automatic checking) can be included into the training data to fine-tune the network. This step can be repeated to progressively improve the training of our network, until no reasonable segmentations of new data can be included. Experimental results demonstrate the effectiveness of our proposed progressive semi-supervised learning fashion as well as its advantage in terms of accuracy.
从磁共振（MR）图像中准确分割盆腔器官在图像引导放射治疗中起着重要作用。然而，由于器官外观不一致和形状变化较大，这是一项具有挑战性的任务。全卷积网络（FCN）最近在医学图像分割中取得了领先的性能，但它需要大量的标记数据进行训练，而在实际情况中通常很难获得这些数据。为了应对这些挑战，我们提出了一种基于深度学习的半监督学习框架。具体来说，我们首先基于有限数量的标记MRI数据训练一个初始的多任务残差全卷积网络（FCN）。基于最初训练的FCN，可以自动分割那些未标记的新数据，并且一些合理的分割结果（经过手动/自动检查）可以包含到训练数据中以微调网络。这一步骤可以重复进行，以逐步改进我们网络的训练，直到无法再包含新数据的合理分割结果为止。实验结果证明了我们提出的渐进式半监督学习方式的有效性及其在准确性方面的优势。
