Exploring convolutional neural networks with transfer learning for diagnosing Lyme disease from skin lesion images

Sk. Imran Hossain, Jocelyn de Goër de Herve, Md Shahriar Hassan, Delphine Martineau, Evelina Petrosyan, Violaine Corbin, Jean Beytout, Isabelle Lebert, Jonas Durand, Irene Carravieri, Annick Brun-Jacob, Pascale Frey-Klett, Elisabeth Baux, Céline Cazorla, Carole Eldin, Yves Hansmann, Solene Patrat-Delon, Thierry Prazuck, Alice Raffetin, Pierre Tattevin et al. (3 additional authors not shown)

Abstract: Highlights•An extensive analysis on a Novel Lyme disease dataset using twenty-three convolutional neural network architectures for Lyme disease diagnosis from images.•Custom transfer learning combining ImageNet and HAM10000 dataset proved effective.•Experimental results suggested even some lightweight convolutional networks are effective for creating Lyme disease pre-scanner mobile applications.•Guideline provided for architecture selection based on different criteria.•Trained models made public which can be used for building Lyme disease pre-scanners and transfer learning.