Synthetic Data Generated from CT Scans for Patient Pose Assessment

Manuel Laufer, Dominik Mairhöfer, Malte Sieren, Hauke Gerdes, Fabio Leal dos Reis, Arpad Bischof, Thomas Käster, Erhardt Barth, Jörg Barkhausen, Thomas Martinetz
Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, PMLR 301:882-895, 2026.

Abstract

An adequate diagnostic quality of radiographs is essential for reliable diagnoses and treatment planning. The patientś pose during radiography is one of the most important factors determining the diagnostic quality. Since patient positioning is difficult and not standardized, an automated AI-based approach using depth images to automatically assess the patientś pose before the radiograph has been taken would be helpful.Due to regulatory hurdles, however, it is difficult in practice to acquire the required depth images and corresponding radiographs.In this paper, we present a framework that can generate such training data synthetically from Computer Tomography scans. We further show that by pretraining on our generated synthetic dataset consisting of 3077 image pairs of upper ankle joints, the pose assessment of real upper ankle joints can be improved by up to 11 percentage points.

Cite this Paper

BibTeX
@InProceedings{pmlr-v301-laufer26a, title = {Synthetic Data Generated from CT Scans for Patient Pose Assessment}, author = {Laufer, Manuel and Mairh\"ofer, Dominik and Sieren, Malte and Gerdes, Hauke and dos Reis, Fabio Leal and Bischof, Arpad and K\"aster, Thomas and Barth, Erhardt and Barkhausen, J\"org and Martinetz, Thomas}, booktitle = {Proceedings of The 8th International Conference on Medical Imaging with Deep Learning}, pages = {882--895}, year = {2026}, editor = {Tasdizen, Tolga and Elhabian, Shireen and Summers, Ronald and Chen, Chen and Koch, Lisa and Zhuang, Yan}, volume = {301}, series = {Proceedings of Machine Learning Research}, month = {09--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v301/main/assets/laufer26a/laufer26a.pdf}, url = {https://proceedings.mlr.press/v301/laufer26a.html}, abstract = {An adequate diagnostic quality of radiographs is essential for reliable diagnoses and treatment planning. The patientś pose during radiography is one of the most important factors determining the diagnostic quality. Since patient positioning is difficult and not standardized, an automated AI-based approach using depth images to automatically assess the patientś pose before the radiograph has been taken would be helpful.Due to regulatory hurdles, however, it is difficult in practice to acquire the required depth images and corresponding radiographs.In this paper, we present a framework that can generate such training data synthetically from Computer Tomography scans. We further show that by pretraining on our generated synthetic dataset consisting of 3077 image pairs of upper ankle joints, the pose assessment of real upper ankle joints can be improved by up to 11 percentage points.} }
Endnote
%0 Conference Paper %T Synthetic Data Generated from CT Scans for Patient Pose Assessment %A Manuel Laufer %A Dominik Mairhöfer %A Malte Sieren %A Hauke Gerdes %A Fabio Leal dos Reis %A Arpad Bischof %A Thomas Käster %A Erhardt Barth %A Jörg Barkhausen %A Thomas Martinetz %B Proceedings of The 8th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Tolga Tasdizen %E Shireen Elhabian %E Ronald Summers %E Chen Chen %E Lisa Koch %E Yan Zhuang %F pmlr-v301-laufer26a %I PMLR %P 882--895 %U https://proceedings.mlr.press/v301/laufer26a.html %V 301 %X An adequate diagnostic quality of radiographs is essential for reliable diagnoses and treatment planning. The patientś pose during radiography is one of the most important factors determining the diagnostic quality. Since patient positioning is difficult and not standardized, an automated AI-based approach using depth images to automatically assess the patientś pose before the radiograph has been taken would be helpful.Due to regulatory hurdles, however, it is difficult in practice to acquire the required depth images and corresponding radiographs.In this paper, we present a framework that can generate such training data synthetically from Computer Tomography scans. We further show that by pretraining on our generated synthetic dataset consisting of 3077 image pairs of upper ankle joints, the pose assessment of real upper ankle joints can be improved by up to 11 percentage points.
APA
Laufer, M., Mairhöfer, D., Sieren, M., Gerdes, H., dos Reis, F.L., Bischof, A., Käster, T., Barth, E., Barkhausen, J. & Martinetz, T.. (2026). Synthetic Data Generated from CT Scans for Patient Pose Assessment. Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 301:882-895 Available from https://proceedings.mlr.press/v301/laufer26a.html.

Related Material