Self-dual pattern spectra for characterising the dermal-epidermal junction in 3D reflectance confocal microscopy imaging

Julie Robic (1, 2) , Benjamin Perret (1) , Alex Nkengne (2) , Michel Couprie (1) , Hugues Talbot (1, 3, 4)

1 LIGM - Laboratoire d'Informatique Gaspard-Monge
2 Laboratoires Clarins
3 CVN - Centre de vision numérique
4 OPIS - OPtimisation Imagerie et Santé

The Dermal-Epidermal Junction (DEJ) is a 2D surface separating the epidermis from the dermis which undergoes multiple changes under pathological or ageing conditions. Recent advances in reflectance confocal microscopy now enables the extraction of the DEJ from in-vivo imaging. This articles proposes a method to automatically analyse DEJ surfaces using self-dual morphological filters. We use self-dual pattern spectra with non-increasing attributes and we propose a novel measure in order to characterize the evolution of the surface under the filtering process. The proposed method is assessed on a specifically constituted dataset and we show that the proposed surface feature significantly correlates with both chronological ageing and photo-ageing.

Keywords: 2D surface; Skin; Tree-of-Shapes; Pattern spectrum

Domains: Image Processing [eess.IV]

Julie Robic, Benjamin Perret, Alex Nkengne, Michel Couprie, Hugues Talbot. Self-dual pattern spectra for characterising the dermal-epidermal junction in 3D reflectance confocal microscopy imaging. ISMM 2019 - International Symposium on Mathematical Morphology and Its Applications to Signal and Image Processing, Jul 2019, Saarbrücken, Germany. pp.508-519, ⟨10.1007/978-3-030-20867-7_39⟩. ⟨hal-02169702⟩