Estimating lumbar bone mineral density from conventional MRI and radiographs with deep learning in spine patients
European Spine Journal · 2024
DOI: 10.1007/s00586-024-08463-8 ↗Kurz gefasst
Fachartikel in der Zeitschrift European Spine Journal, 2024. Verfasst von Fabio Galbusera und 8 weiteren Autor:innen. Beteiligte Einrichtung: Department of Teaching, Research and Development, Schulthess Clinic, Lengghalde 2, Zurich, 8008, Switzerland. fabio.galbusera@kws.ch. Hauptthemen laut MEDLINE: Bone Density, Deep Learning, Lumbar Vertebrae, Magnetic Resonance Imaging. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 39 Literaturangaben. Gefördert von: International Society for the Study of the Lumbar Spine.
Abstract (Originalsprache)
PURPOSE: This study aimed to develop machine learning methods to estimate bone mineral density and detect osteopenia/osteoporosis from conventional lumbar MRI (T1-weighted and T2-weighted images) and planar radiography in combination with clinical data and imaging parameters of the acquisition protocol. METHODS: A database of 429 patients subjected to lumbar MRI, radiographs and dual-energy x-ray absorptiometry within 6 months was created from an institutional database. Several machine learning models were trained and tested (373 patients for training, 86 for testing) with the following objectives: (1) direct estimation of the vertebral bone mineral density; (2) classification of T-score lower than - 1 or (3) lower than - 2.5. The models took as inputs either the images or radiomics features derived from them, alone or in combination with metadata (age, sex, body size, vertebral level, parameters of the imaging protocol). RESULTS: The best-performing models achieved mean absolute errors of 0.15-0.16 g/cm2 for the direct estimation of bone mineral density, and areas under the receiver operating characteristic curve of 0.82 (MRIs) - 0.80 (radiographs) for the classification of T-scores lower than - 1, and 0.80 (MRIs) - 0.65 (radiographs) for T-scores lower than - 2.5. CONCLUSIONS: The models showed good discriminative performances in detecting cases of low bone mineral density, and more limited capabilities for the direct estimation of its value. Being based on routine imaging and readily available data, such models are promising tools to retrospectively analyse existing datasets as well as for the opportunistic investigation of bone disorders.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Teaching, Research and Development, Schulthess Clinic, Lengghalde 2, Zurich, 8008, Switzerland. fabio.galbusera@kws.ch.
- Department of Teaching, Research and Development, Schulthess Clinic, Lengghalde 2, Zurich, 8008, Switzerland.
- Department of Health Sciences and Technology (D-HEST), ETH Zürich, Zurich, Switzerland.
ORCID-Profile: Fabio Galbusera, Andrea Cina, Jacopo A Vitale, Markus Loibl, Tamás F Fekete, Frank Kleinstück, Daniel Haschtmann, Anne F Mannion
Förderung
- International Society for the Study of the Lumbar Spine
Zitieren
APA
Galbusera, F., Cina, A., O'Riordan, D., Vitale, J. A., Loibl, M., Fekete, T. F., Kleinstück, F., Haschtmann, D., & Mannion, A. F. (2024). Estimating lumbar bone mineral density from conventional MRI and radiographs with deep learning in spine patients. European Spine Journal. https://doi.org/10.1007/s00586-024-08463-8
Vancouver
Galbusera F, Cina A, O'Riordan D, Vitale JA, Loibl M, Fekete TF, et al. Estimating lumbar bone mineral density from conventional MRI and radiographs with deep learning in spine patients. European Spine Journal. 2024. doi:10.1007/s00586-024-08463-8
BibTeX
@article{Galbusera2024,
author = {Galbusera, Fabio and Cina, Andrea and O'Riordan, Dave and Vitale, Jacopo A and Loibl, Markus and Fekete, Tamás F and Kleinstück, Frank and Haschtmann, Daniel and Mannion, Anne F},
title = {Estimating lumbar bone mineral density from conventional MRI and radiographs with deep learning in spine patients},
journal = {European Spine Journal},
year = {2024},
doi = {10.1007/s00586-024-08463-8}
}
EUVIMED