Combining neural networks and image synthesis to enable automatic thoracic cavity segmentation of hyperpolarized...

Quantification of ¹²⁹Xe MRI relies on accurate segmentation of the thoracic cavity, typically performed manually using a combination of ¹H and ¹²⁹Xe scans. This can be accelerated by using Convolutional Neural Networks (CNNs) that segment only the ¹²⁹Xe scan. However, this task is complicated by peripheral ventilation defects, which requires training CNNs with large, diverse datasets. Here, we accelerate the creation of training data by synthesizing ¹²⁹Xe images with a variety of defects. We use this to train a 3D model to provide thoracic cavity segmentation from ¹²⁹Xe ventilation MRI alone.

Training and testing data consisted of 22 and 33 3D ¹²⁹Xe ventilation images. Training data were expanded to 484 using Template-based augmentation while an additional 298 images were synthesized using the Pix2Pix model. This data was used to train both a 2D U-net and 3D V-net-based segmentation model using a combination of Dice-Focal and Anatomical Constraint loss functions. Segmentation performance was compared using Dice coefficients calculated over the entire lung and within ventilation defects.

Performance of both U-net and 3D segmentation was improved by including synthetic training data. The 3D models performed significantly better than U-net, and the 3D model trained with synthetic ¹²⁹Xe images exhibited the highest overall Dice score of 0.929. Moreover, addition of synthetic training data improved the Dice score in ventilation defect regions from 0.545 to 0.588 for U-net and 0.739 to 0.765 for the 3D model.

It is feasible to obtain high-quality segmentations from ¹²⁹Xe scan alone using 3D models trained with additional synthetic images.