Locating Cephalometric X-Ray Landmarks with Foveated Pyramid Attention

Logan Gilmour, Nilanjan Ray
Proceedings of the Third Conference on Medical Imaging with Deep Learning, PMLR 121:262-276, 2020.

Abstract

CNNs, initially inspired by human vision, differ in a key way: they sample uniformly, rather than with highest density in a focal point. For very large images, this makes training untenable, as the memory and computation required for activation maps scales quadratically with the side length of an image. We propose an image pyramid based approach that extracts narrow glimpses of the of the input image and iteratively refines them to accomplish regression tasks. To assist with high-accuracy regression, we introduce a novel intermediate representation we call ‘spatialized features’. Our approach scales logarithmically with the side length, so it works with very large images. We apply our method to Cephalometric X-ray Landmark Detection and get state-of-the-art results.

Cite this Paper

BibTeX
@InProceedings{pmlr-v121-gilmour20a, title = {Locating Cephalometric X-Ray Landmarks with Foveated Pyramid Attention}, author = {Gilmour, Logan and Ray, Nilanjan}, booktitle = {Proceedings of the Third Conference on Medical Imaging with Deep Learning}, pages = {262--276}, year = {2020}, editor = {Arbel, Tal and Ben Ayed, Ismail and de Bruijne, Marleen and Descoteaux, Maxime and Lombaert, Herve and Pal, Christopher}, volume = {121}, series = {Proceedings of Machine Learning Research}, month = {06--08 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v121/gilmour20a/gilmour20a.pdf}, url = {https://proceedings.mlr.press/v121/gilmour20a.html}, abstract = {CNNs, initially inspired by human vision, differ in a key way: they sample uniformly, rather than with highest density in a focal point. For very large images, this makes training untenable, as the memory and computation required for activation maps scales quadratically with the side length of an image. We propose an image pyramid based approach that extracts narrow glimpses of the of the input image and iteratively refines them to accomplish regression tasks. To assist with high-accuracy regression, we introduce a novel intermediate representation we call ‘spatialized features’. Our approach scales logarithmically with the side length, so it works with very large images. We apply our method to Cephalometric X-ray Landmark Detection and get state-of-the-art results.} }
Endnote
%0 Conference Paper %T Locating Cephalometric X-Ray Landmarks with Foveated Pyramid Attention %A Logan Gilmour %A Nilanjan Ray %B Proceedings of the Third Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2020 %E Tal Arbel %E Ismail Ben Ayed %E Marleen de Bruijne %E Maxime Descoteaux %E Herve Lombaert %E Christopher Pal %F pmlr-v121-gilmour20a %I PMLR %P 262--276 %U https://proceedings.mlr.press/v121/gilmour20a.html %V 121 %X CNNs, initially inspired by human vision, differ in a key way: they sample uniformly, rather than with highest density in a focal point. For very large images, this makes training untenable, as the memory and computation required for activation maps scales quadratically with the side length of an image. We propose an image pyramid based approach that extracts narrow glimpses of the of the input image and iteratively refines them to accomplish regression tasks. To assist with high-accuracy regression, we introduce a novel intermediate representation we call ‘spatialized features’. Our approach scales logarithmically with the side length, so it works with very large images. We apply our method to Cephalometric X-ray Landmark Detection and get state-of-the-art results.
APA
Gilmour, L. & Ray, N.. (2020). Locating Cephalometric X-Ray Landmarks with Foveated Pyramid Attention. Proceedings of the Third Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 121:262-276 Available from https://proceedings.mlr.press/v121/gilmour20a.html.

Related Material