High Resolution Human Skin Image Segmentation by means of Fully Convolutional Neural Networks

• Published in: 2018 International Conference on Numerical Simulation of Optoelectronic Devices (NUSOD) pp. 31 - 32
• Main Authors: Calderon-Delgado, Manuel, Tjiu, Jeng-Wei, Lin, Ming-Yi, Huang, Sheng-Lung
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.11.2018
• Subjects: convolutional neural networks; Dermis; Epidermis; Glass; Image resolution; Image segmentation; machine learning; medical imaging; optical coherence tomography; skin; Training

Convolutional neural networks were applied to sub-micron-resolution optical coherence tomography images of the human skin for anatomical segmentation. The main layers of skin were discerned with an average 90% accuracy, which we believe to possess potential in the assessment of skin health.