BACKGROUND: Pseudoxanthoma elasticum (PXE) is a hereditary disease that manifests - among others - with papular lesions on the skin. The presence of these papules is the earliest sign of the pathology. Their detection is therefore of importance but, due to their small size, this can become a difficult task. This is why dermoscopy, a noninvasive imaging modality, may be of interest. Nevertheless, the detection of papules on dermoscopic images is still a challenge. OBJECTIVE: We propose an image processing framework to help in the detection of papules from dermoscopic images. Our algorithm is based on the recently-proposed bi-dimensional sample entropy. METHODS: Seven PXE patients participated to the study. For each of them, one dermoscopic image of the neck (where the papules are predominant), and one dermoscopic image of a normal skin zone have been recorded. For each one, the bi-dimensional sample entropy has been computed. RESULTS: We observed statistically significantly lower bi-dimensional sample entropy values on the dermoscopic images of the neck than on the dermoscopic images of the normal zone. CONCLUSION: These preliminary findings show that the bi-dimensional sample entropy might be of interest for the diagnosis and follow-up of the PXE pathology.