September 5, 2025Scientific ReportsOpen Access

Comparative analysis of supervised and self-supervised learning with small and imbalanced medical imaging datasets

View Full Paper

Ask AI

Bookmark

Authors

AEAndrea Espis CMChiara MarziUniversity of Florence SDStefano Diciotti

Discussion

Member takes

Implication

This analysis compares self-supervised learning and supervised learning in small, imbalanced medical imaging datasets, indicating performance differences.

Key Points

In most scenarios, supervised learning outperformed self-supervised learning on small, imbalanced datasets.
Four binary classification tasks included age prediction and diagnosis of Alzheimer's disease using MRI scans.
Analyzed training sets of various sizes, with the mean size ranging from 843 to 33,484 images across tasks.
Findings suggest that selection of learning paradigms should be tailored based on dataset characteristics.

Cite This Study

Espis et al. (2025) studied this question.

synapsesocial.com/papers/68c23a74b210217d647804db https://doi.org/10.1038/s41598-025-99000-0

View Full Paper

Ask AI

Bookmark

Implication

This analysis compares self-supervised learning and supervised learning in small, imbalanced medical imaging datasets, indicating performance differences.

Comparative analysis of supervised and self-supervised learning with small and imbalanced medical imaging datasets

Discussion

Member takes

Implication

Key Points

Cite This Study

Comparative analysis of supervised and self-supervised learning with small and imbalanced medical imaging datasets

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider