Fully Convolutional Architectures for Multi-Class Segmentation in Chest Radiographs
This work addresses the problem of accurate anatomical segmentation in medical imaging for clinicians, though it is incremental as it adapts existing concepts to a specific domain.
The paper tackled automated multi-class segmentation of lungs, clavicles, and heart in chest radiographs, achieving mean Jaccard scores of 95.0% for lungs, 86.8% for clavicles, and 88.2% for heart, outperforming state-of-the-art methods and human observers on lungs and heart.
The success of deep convolutional neural networks on image classification and recognition tasks has led to new applications in very diversified contexts, including the field of medical imaging. In this paper we investigate and propose neural network architectures for automated multi-class segmentation of anatomical organs in chest radiographs, namely for lungs, clavicles and heart. We address several open challenges including model overfitting, reducing number of parameters and handling of severely imbalanced data in CXR by fusing recent concepts in convolutional networks and adapting them to the segmentation problem task in CXR. We demonstrate that our architecture combining delayed subsampling, exponential linear units, highly restrictive regularization and a large number of high resolution low level abstract features outperforms state-of-the-art methods on all considered organs, as well as the human observer on lungs and heart. The models use a multi-class configuration with three target classes and are trained and tested on the publicly available JSRT database, consisting of 247 X-ray images the ground-truth masks for which are available in the SCR database. Our best performing model, trained with the loss function based on the Dice coefficient, reached mean Jaccard overlap scores of 95.0\% for lungs, 86.8\% for clavicles and 88.2\% for heart. This architecture outperformed the human observer results for lungs and heart.