Boosting EfficientNets Ensemble Performance via Pseudo-Labels and Synthetic Images by pix2pixHD for Infection and Ischaemia Classification in Diabetic Foot Ulcers

TL;DR

This paper demonstrates that augmenting training data with pseudo-labels and synthetic images generated by pix2pixHD significantly improves the classification of infection and ischaemia in diabetic foot ulcers, especially for rare classes.

Contribution

The study introduces a novel data augmentation approach combining pseudo-labeling and synthetic image generation to enhance EfficientNet ensemble performance in medical image classification.

Findings

Enlarged training dataset increased by 8.68 times with synthetic images.

Models trained on extended data show significant F1-score improvements for rare classes.

Synthetic images provide broad qualitative variety, aiding model generalization.

Abstract

Diabetic foot ulcers are a common manifestation of lesions on the diabetic foot, a syndrome acquired as a long-term complication of diabetes mellitus. Accompanying neuropathy and vascular damage promote acquisition of pressure injuries and tissue death due to ischaemia. Affected areas are prone to infections, hindering the healing progress. The research at hand investigates an approach on classification of infection and ischaemia, conducted as part of the Diabetic Foot Ulcer Challenge (DFUC) 2021. Different models of the EfficientNet family are utilized in ensembles. An extension strategy for the training data is applied, involving pseudo-labeling for unlabeled images, and extensive generation of synthetic images via pix2pixHD to cope with severe class imbalances. The resulting extended training dataset features $8.68$ times the size of the baseline and shows a real to synthetic image ratio of $1:3$. Performances of models and ensembles trained on the baseline and extended training dataset are compared. Synthetic images featured a broad qualitative variety. Results show that models trained on the extended training dataset as well as their ensemble benefit from the large extension. F1-Scores for rare classes receive outstanding boosts, while those for common classes are either not harmed or boosted moderately. A critical discussion concretizes benefits and identifies limitations, suggesting improvements. The work concludes that classification performance of individual models as well as that of ensembles can be boosted utilizing synthetic images. Especially performance for rare classes benefits notably.