Kordon F, Maier A, Kunze H (2021)
Publication Language: English
Publication Type: Conference contribution, Conference Contribution
Publication year: 2021
Publisher: Springer Vieweg
City/Town: Wiesbaden
Pages Range: 350-355
Conference Proceedings Title: Bildverarbeitung für die Medizin 2021
Event location: Virtual Conference
DOI: 10.1007/978-3-658-33198-6_85
Vertebral corner points are frequently used landmarks for a vast variety of orthopedic and trauma surgical applications. Algorithmic approaches that are designed to automatically detect them on 2D radiographs have to cope with varying image contrast, high noise levels, and superimposed soft tissue. To enforce an anatomically correct landmark configuration in presence of these limitations, this study investigates a shape constraint technique based on data-driven encodings of the spine geometry. A contractive PointNet autoencoder is used to map numerical landmark coordinate representations onto a low-dimensional shape manifold. A distance norm between prediction and ground truth encodings then serves as an additional loss term during optimization. The method is compared and evaluated on the SpineWeb16 dataset. Small improvements can be observed, recommending further analysis of the encoding design and composite cost function.
APA:
Kordon, F., Maier, A., & Kunze, H. (2021). Latent Shape Constraint for Anatomical Landmark Detection on Spine Radiographs. In Palm C, Deserno TM, Handels H, Maier A, Maier-Hein K, Tolxdorff T (Eds.), Bildverarbeitung für die Medizin 2021 (pp. 350-355). Virtual Conference: Wiesbaden: Springer Vieweg.
MLA:
Kordon, Florian, Andreas Maier, and Holger Kunze. "Latent Shape Constraint for Anatomical Landmark Detection on Spine Radiographs." Proceedings of the Bildverarbeitung für die Medizin 2021, Virtual Conference Ed. Palm C, Deserno TM, Handels H, Maier A, Maier-Hein K, Tolxdorff T, Wiesbaden: Springer Vieweg, 2021. 350-355.
BibTeX: Download