Comparison of Image Normalization Methods for Multi-Site Deep Learning
Applied Sciences · 03.08.2023 · 13 (15) · 8923
DOI: 10.3390/app13158923 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 03.08.2023. Verfasst von Steffen Albert und 7 weiteren Autor:innen. Die Arbeit wurde bisher 22-mal zitiert (Crossref) und nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: German Research Foundation (DFG), Ministry of Science, Research, Arts Baden-Württemberg (MWK).
Abstract (Originalsprache)
In this study, we evaluate the influence of normalization on the performance of deep learning networks for tumor segmentation and the prediction of the pathological response of locally advanced rectal cancer to neoadjuvant chemoradiotherapy. The techniques were applied to a multicenter and multimodal magnet resonance imaging data set consisting of 201 patients recorded at six centers. We implemented and investigated six different normalization methods (setting the mean and standard deviation, histogram matching, percentiles, combining percentiles and histogram matching, fixed window and an auto-encoder with adversarial loss using the imaging parameters) and evaluated their impact on four deep learning tasks: tumor segmentation, prediction of treatment outcome, and prediction of sex and age. The latter two tasks were implemented as a reference test. We trained a modified U-Net with different normalization methods in multiple configurations: on all images, images from all centers except one, and images from a single center. Our results show that normalization only plays a minor role in segmentation, with a difference in Dice of less than 0.02 between the best and worst performing networks. For the prediction of sex and treatment outcomes, the percentile method combined with histogram matching works best for all scenarios. The biggest difference in performance, depending on the normalization method, occurs for classification. In conclusion, normalization is especially important for small data sets or for generalizing to different data distributions. The deep learning method was superior to the classical methods only in a minority of cases, probably due to the limited amount of training data.
Förderung
- German Research Foundation (DFG)
- Ministry of Science, Research
- Arts Baden-Württemberg (MWK)
Zitieren
APA
Albert, S., Wichtmann, B. D., Zhao, W., Maurer, A., Hesser, J., Attenberger, U. I., Schad, L. R., & Zöllner, F. G. (2023). Comparison of Image Normalization Methods for Multi-Site Deep Learning. Applied Sciences, 13(15), 8923. https://doi.org/10.3390/app13158923
Vancouver
Albert S, Wichtmann BD, Zhao W, Maurer A, Hesser J, Attenberger UI, et al. Comparison of Image Normalization Methods for Multi-Site Deep Learning. Applied Sciences. 2023;13(15):8923. doi:10.3390/app13158923
BibTeX
@article{Albert2023,
author = {Albert, Steffen and Wichtmann, Barbara D. and Zhao, Wenzhao and Maurer, Angelika and Hesser, Jürgen and Attenberger, Ulrike I. and Schad, Lothar R. and Zöllner, Frank G.},
title = {Comparison of Image Normalization Methods for Multi-Site Deep Learning},
journal = {Applied Sciences},
year = {2023},
volume = {13},
number = {15},
pages = {8923},
doi = {10.3390/app13158923}
}
EUVIMED