Activities per year
Abstract
In many medical image analysis applications, only a limited amount of training data is available due to the costs of image acquisition and the large manual annotation effort required from experts. Training recent state-of-the-art machine learning methods like convolutional neural networks (CNNs) from small datasets is a challenging task. In this work on anatomical landmark localization, we propose a CNN architecture that learns to split the localization task into two simpler sub-problems, reducing the overall need for large training datasets. Our fully convolutional SpatialConfiguration-Net (SCN) learns this simplification due to multiplying the heatmap predictions of its two components and by training the network in an end-to-end manner. Thus, the SCN dedicates one component to locally accurate but ambiguous candidate predictions, while the other component improves robustness to ambiguities by incorporating the …
Original language | English |
---|---|
Pages (from-to) | 207-219 |
Journal | Medical Image Analysis |
Volume | 54 |
DOIs | |
Publication status | Published - 2019 |
-
Integrating Spatial Configuration into Heatmap Regression Based CNNs for Landmark Localization
Christian Payer (Speaker)
2019Activity: Talk or presentation › Poster presentation › Science to science
-
Integrating Spatial Configuration into CNNs for Anatomical Landmark Localization and Multi-Label Whole Heart Segmentation
Christian Payer (Speaker)
2019Activity: Talk or presentation › Invited talk at conference or symposium › Science to science