Patient Pose Assessment Using a CT-Based Framework for Synthetic Data Generation

Manuel Laufer1, Dominik Mairhöfer1, Malte Sieren2, Hauke Gerdes2, Fabio Leal dos Reis2, Arpad Bischof2,3, Thomas Käster4, Erhardt Barth1, Jörg Barkhausen2, Thomas Martinetz1
1: Institute for Neuro- and Bioinformatics, University of Lübeck, Germany, 2: University Medical Center Schleswig-Holstein, Lübeck, Germany, 3: IMAGE Information Systems Europe GmbH, Rostock, Germany, 4: Pattern Recognition Company GmbH, Lübeck, Germany
Publication date: 2026/08/28
https://doi.org/10.59275/j.melba.2026-c874
PDF · Synthetic dataset

Abstract

An adequate diagnostic quality of radiographs is essential for reliable diagnoses and treatment planning. The patient’s 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 patient’s 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 Computed 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.

Keywords

Patient Pose Assessment · Synthetic Data Generation · Diagnostic Quality · CT Scan · Time-of-Flight Cameras · Radiography · Deep Learning

Bibtex @article{melba:2026:027:laufer, title = "Patient Pose Assessment Using a CT-Based Framework for Synthetic Data Generation", author = "Laufer, Manuel and Mairhöfer, Dominik and Sieren, Malte and Gerdes, Hauke and Leal dos Reis, Fabio and Bischof, Arpad and Käster, Thomas and Barth, Erhardt and Barkhausen, Jörg and Martinetz, Thomas", journal = "Machine Learning for Biomedical Imaging", volume = "2026", issue = "MIDL 2025 special issue", year = "2026", pages = "553--570", issn = "2766-905X", doi = "https://doi.org/10.59275/j.melba.2026-c874", url = "https://melba-journal.org/2026:027" }
RISTY - JOUR AU - Laufer, Manuel AU - Mairhöfer, Dominik AU - Sieren, Malte AU - Gerdes, Hauke AU - Leal dos Reis, Fabio AU - Bischof, Arpad AU - Käster, Thomas AU - Barth, Erhardt AU - Barkhausen, Jörg AU - Martinetz, Thomas PY - 2026 TI - Patient Pose Assessment Using a CT-Based Framework for Synthetic Data Generation T2 - Machine Learning for Biomedical Imaging VL - 2026 IS - MIDL 2025 special issue SP - 553 EP - 570 SN - 2766-905X DO - https://doi.org/10.59275/j.melba.2026-c874 UR - https://melba-journal.org/2026:027 ER -
2026:027 cover