Fine-tuning pre-trained neural networks for medical image classification in small clinical datasets

Funding We would like to acknowledge eurekaSD: Enhancing University Research and Education in Areas Useful for Sustainable Development - grants EK14AC0037 and EK15AC0264. We thank Araucária Foundation for the Support of the Scientific and Technological Development of Paraná through a Research and Te...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Multimedia tools and applications 2024-03, Vol.83 (9), p.27305-27329
Hauptverfasser: Spolaôr, Newton, Lee, Huei Diana, Mendes, Ana Isabel, Nogueira, Conceição, Parmezan, Antonio Rafael Sabino, Takaki, Weber Shoity Resende, Coy, Claudio Saddy Rodrigues, Wu, Feng Chung, Fonseca-Pinto, Rui
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Funding We would like to acknowledge eurekaSD: Enhancing University Research and Education in Areas Useful for Sustainable Development - grants EK14AC0037 and EK15AC0264. We thank Araucária Foundation for the Support of the Scientific and Technological Development of Paraná through a Research and Technological Productivity Scholarship for H. D. Lee (grant 028/2019). We also thank the Brazilian National Council for Scientific and Technological Development (CNPq) through the grant number 142050/2019-9 for A. R. S. Parmezan. The Portuguese team was partially supported by Fundação para a Ciência e a Tecnologia (FCT). R. Fonseca-Pinto was financed by the projects UIDB/50008/2020, UIDP/50008/2020, UIDB/05704/2020 and UIDP/05704/2020 and C. V. Nogueira was financed by the projects UIDB/00013/2020 and UIDP/00013/2020. The funding agencies did not have any further involvement in this paper. Convolutional neural networks have been effective in several applications, arising as a promising supporting tool in a relevant Dermatology problem: skin cancer diagnosis. However, generalizing well can be difficult when little training data is available. The fine-tuning transfer learning strategy has been employed to differentiate properly malignant from non-malignant lesions in dermoscopic images. Fine-tuning a pre-trained network allows one to classify data in the target domain, occasionally with few images, using knowledge acquired in another domain. This work proposes eight fine-tuning settings based on convolutional networks previously trained on ImageNet that can be employed mainly in limited data samples to reduce overfitting risk. They differ on the architecture, the learning rate and the number of unfrozen layer blocks. We evaluated the settings in two public datasets with 104 and 200 dermoscopic images. By finding competitive configurations in small datasets, this paper illustrates that deep learning can be effective if one has only a few dozen malignant and non-malignant lesion images to study and differentiate in Dermatology. The proposal is also flexible and potentially useful for other domains. In fact, it performed satisfactorily in an assessment conducted in a larger dataset with 746 computerized tomographic images associated with the coronavirus disease.
ISSN:1573-7721
1380-7501
1573-7721
DOI:10.1007/s11042-023-16529-w