Data for: Deep Learning Approach to Skin Layers Segmentation in Inflammatory Dermatoses
Published: 20 August 2021| Version 1 | DOI: 10.17632/5p7fxjt7vs.1
Contributors:
, , , Description
The dataset consists of 380 High Frequency Ultrasound (HFUS) images of 380 different patients with AD (303) and psoriasis (77). The data were acquired using DUB SkinScanner75 with a 75 MHz transducer yielding images of size 2067x1555 pix and of four different resolutions (lateral x axial): {0.0019x0.085, 0.0024x0.085, 0.0031x0.085, 0.0019x0.085} mm/pix. For each image, the entry echo - epidermis and SLEB layer delineated by an expert are provided. The data set includes the neural network models trained to segment the layers. The details concerning the models are given in reference paper: https://doi.org/10.1016/j.ultras.2021.106412
Files
Institutions
Politechnika Slaska
Categories
Radiology, Biomedical Engineering, Skin Disease