KEYWORDS: Education and training, Image segmentation, Overfitting, Interpolation, 3D modeling, Volume rendering, Medical imaging, Data modeling, Active learning, 3D applications
PurposeSemantic segmentation is one of the most significant tasks in medical image computing, whereby deep neural networks have shown great success. Unfortunately, supervised approaches are very data-intensive, and obtaining reliable annotations is time-consuming and expensive. Sparsely labeled approaches, such as bounding boxes, have shown some success in reducing the annotation time. However, in 3D volume data, each slice must still be manually labeled.ApproachWe evaluate approaches that reduce the annotation effort by reducing the number of slices that need to be labeled in a 3D volume. In a two-step process, a similarity metric is used to select slices that should be annotated by a trained radiologist. In the second step, a predictor is used to predict the segmentation mask for the rest of the slices. We evaluate different combinations of selectors and predictors on medical CT and MRI volumes. Thus we can determine that combination works best, and how far slice annotations can be reduced.ResultsOur results show that for instance for the Medical Segmentation Decathlon—heart dataset, some selector, and predictor combinations allow for a Dice score 0.969 when only annotating 20% of slices per volume. Experiments on other datasets show a similarly positive trend.ConclusionsWe evaluate a method that supports experts during the labeling of 3D medical volumes. Our approach makes it possible to drastically reduce the number of slices that need to be manually labeled. We present a recommendation in which selector predictor combination to use for different tasks and goals.
We compare axial 2D U-Nets and their 3D counterparts for pixel/voxel-based segmentation of five abdominal organs in CT scans. For each organ, two competing CNNs are trained. They are evaluated by performing five-fold cross-validation on 80 3D images. In a two-step concept, the relevant area containing the organ is first extracted by detected bounding boxes and then passed as input to the organ-specific U-Net. Furthermore, a random regression forest approach for the automatic detection of bounding boxes is summarized from our previous work. The results show that the 2D U-Net is mostly on par with the 3D U-Net or even outperforms it. Especially for the kidneys, it is significantly better suited in this study.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.