DermoGAN: multi-task cycle generative adversarial networks for unsupervised automatic cell identification on in-vivo reflectance confocal microscopy images of the human epidermis - Archive ouverte HAL
Article Dans Une Revue Journal of Biomedical Optics Année : 2024

DermoGAN: multi-task cycle generative adversarial networks for unsupervised automatic cell identification on in-vivo reflectance confocal microscopy images of the human epidermis

Résumé

Significance: Accurate identification of epidermal cells on reflectance confocal microscopy (RCM) images is important in the study of epidermal architecture and topology of both healthy and diseased skin. However, analysis of these images is currently done manually, and therefore time-consuming, subject to human error and inter-expert interpretation. It is also hindered by low image quality due to noise and heterogeneity. Aim: To design an automated pipeline for the analysis of the epidermal structure from reflectance confocal microscopy images. Approach: Two attempts have been made at automatically localizing epidermal cells, called keratinocytes, on RCM images, the first based on a rotationally symmetric error function mask, and the second on cell morphological features. Here we propose a dual-task network to automatically identify keratinocytes on RCM images. Each task consists of a cycle Generative Adversarial Network (cycle-GAN). The first task aims to translate real RCM images into binary images thus learning the noise and texture model of RCM images, whereas the second task maps Gabor-filtered RCM images into binary images, learning the epidermal structure visible on RCM images. The combination of the two tasks allows one task to constrict the solution space of the other thus improving overall results. We refine our cell identification by applying the pre-trained StarDist algorithm to detect star-convex shapes, thus closing any incomplete membranes and separating neighboring cells. Results: The results are evaluated both on simulated data and on manually annotated real RCM data. Accuracy is measured using recall and precision metrics, which as summarized as the F1-score. Conclusions: We demonstrate that the proposed fully unsupervised method successfully identifies keratinocytes on RCM images of the epidermis, with an accuracy on par with experts’ cell identification and is not constrained by limited annotated available data and can be extended to images acquired using various imaging techniques with no-retraining. Keywords: cycle-GAN, identification, keratinocytes, multi-task, reflectance confocal microscopy, segmentation.
Fichier principal
Vignette du fichier
DermoGAN_paper.pdf (9.38 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04670338 , version 1 (12-08-2024)

Identifiants

Citer

Imane Lboukili, Georgios Stamatas, Xavier Descombes. DermoGAN: multi-task cycle generative adversarial networks for unsupervised automatic cell identification on in-vivo reflectance confocal microscopy images of the human epidermis. Journal of Biomedical Optics, 2024, 29 (08), pp.086003. ⟨10.1117/1.JBO.29.8.086003⟩. ⟨hal-04670338⟩
59 Consultations
14 Téléchargements

Altmetric

Partager

More