Towards Reducing the Need for Annotations in Digital Dermatology with Self-supervised Learning

Fabian Gröger, Philippe Gottfrois, Ludovic Amruthalingam, Alvaro Gonzalez-Jimenez, Simone Lionetti, Alexander Navarini, Alexander Navarini, Marc Pouly

2022

Abstract

Training supervised models requires large amounts of labelled data, whose creation is often expensive and time-consuming, especially in the medical domain. The standard practice to mitigate the lack of annotated clinical images is to use transfer learning and fine-tune pre-trained ImageNet weights on a downstream task. While this approach achieves satisfactory performance, it still requires a sufficiently large dataset to adjust the global features for a specific task. We report on an ongoing investigation to determine whether self-supervised learning methods applied to unlabelled domain-specific images can provide better representations for digital dermatology compared to ImageNet. We consider ColorMe, SimCLR, BYOL, DINO, and iBOT, and present preliminary results on the evaluation of pre-trained initialization for three different medical tasks with mixed imaging modalities. Our intermediate findings indicate a benefit in using features learned by iBOT on dermatology datasets compared to conventional transfer learning from ImageNet classification.

Paper Citation

in Harvard Style

Gröger F., Gottfrois P., Amruthalingam L., Gonzalez-Jimenez A., Lionetti S., Navarini A. and Pouly M. (2022). Towards Reducing the Need for Annotations in Digital Dermatology with Self-supervised Learning. In Proceedings of the 1st Workshop on Scarce Data in Artificial Intelligence for Healthcare - Volume 1: SDAIH, ISBN 978-989-758-629-3, SciTePress, pages 41-46. DOI: 10.5220/0011532100003523

in Bibtex Style

@conference{sdaih22,
author={Fabian Gröger and Philippe Gottfrois and Ludovic Amruthalingam and Alvaro Gonzalez-Jimenez and Simone Lionetti and Alexander Navarini and Marc Pouly},
title={Towards Reducing the Need for Annotations in Digital Dermatology with Self-supervised Learning},
booktitle={Proceedings of the 1st Workshop on Scarce Data in Artificial Intelligence for Healthcare - Volume 1: SDAIH,},
year={2022},
pages={41-46},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011532100003523},
isbn={978-989-758-629-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st Workshop on Scarce Data in Artificial Intelligence for Healthcare - Volume 1: SDAIH,
TI - Towards Reducing the Need for Annotations in Digital Dermatology with Self-supervised Learning
SN - 978-989-758-629-3
AU - Gröger F.
AU - Gottfrois P.
AU - Amruthalingam L.
AU - Gonzalez-Jimenez A.
AU - Lionetti S.
AU - Navarini A.
AU - Pouly M.
PY - 2022
SP - 41
EP - 46
DO - 10.5220/0011532100003523
PB - SciTePress