Can Synthetic Images Improve CNN Performance in Wound Image Classification?

Malihi L, Hübner U, Richter ML, Moelleken M, Przysucha M, Busch D, Heggemann J, Hafer G, Wiemeyer S, Heidemann G, Dissemond J, Erfurt-Berge C, Barkhau C, Hendriks A, Hüsers J (2023)

Publication Type: Book chapter / Article in edited volumes

Publication year: 2023

Publisher: IOS Press

Edited Volumes: Caring is Sharing – Exploiting the Value in Data for Health and Innovation

Series: Studies in Health Technology and Informatics

Book Volume: 302

Pages Range: 927-931

DOI: 10.3233/SHTI230311

Abstract

For artificial intelligence (AI) based systems to become clinically relevant, they must perform well. Machine Learning (ML) based AI systems require a large amount of labelled training data to achieve this level. In cases of a shortage of such large amounts, Generative Adversarial Networks (GAN) are a standard tool for synthesising artificial training images that can be used to augment the data set. We investigated the quality of synthetic wound images regarding two aspects: (i) improvement of wound-type classification by a Convolutional Neural Network (CNN) and (ii) how realistic such images look to clinical experts (n = 217). Concerning (i), results show a slight classification improvement. However, the connection between classification performance and the size of the artificial data set is still unclear. Regarding (ii), although the GAN could produce highly realistic images, the clinical experts took them for real in only 31% of the cases. It can be concluded that image quality may play a more significant role than data size in improving the CNN-based classification result.

Authors with CRIS profile

Dorothee Busch Department of Dermatology Cornelia Erfurt-Berge Department of Dermatology

Involved external institutions

Universitätsklinikum Essen DE Germany (DE) Universität Osnabrück DE Germany (DE) Symbic GmbH DE Germany (DE) Niels-Stensen-Kliniken GmbH DE Germany (DE)

How to cite

APA:

Malihi, L., Hübner, U., Richter, M.L., Moelleken, M., Przysucha, M., Busch, D.,... Hüsers, J. (2023). Can Synthetic Images Improve CNN Performance in Wound Image Classification? In Maria Hägglund, Madeleine Blusi, Stefano Bonacina, Lina Nilsson, Inge Cort Madsen, Sylvia Pelayo, Anne Moen, Arriel Benis, Lars Lindsköld, Parisis Gallos (Eds.), Caring is Sharing – Exploiting the Value in Data for Health and Innovation. (pp. 927-931). IOS Press.

MLA:

Malihi, Leila, et al. "Can Synthetic Images Improve CNN Performance in Wound Image Classification?" Caring is Sharing – Exploiting the Value in Data for Health and Innovation. Ed. Maria Hägglund, Madeleine Blusi, Stefano Bonacina, Lina Nilsson, Inge Cort Madsen, Sylvia Pelayo, Anne Moen, Arriel Benis, Lars Lindsköld, Parisis Gallos, IOS Press, 2023. 927-931.

BibTeX: Download