Skip to main navigation Skip to search Skip to main content

Artificial Intelligence Based Segmentation for Skin-Layer Visualization from Optical Coherence Tomography Images

  • Gilang Titah Ramadhan
  • , Yori Pusparani
  • , Ardha Ardea Prisilla
  • , Wei Cheng Shen
  • , Hsu Tang Cheng
  • , Ben Yi Liau
  • , Yih Kuen Jan
  • , Wen Thong Chang
  • , Chih Yang Lin
  • , Chi Wen Lung

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Changes in epidermal thickness are linked to various skin diseases, such as diabetic foot. Optical Coherence Tomography (OCT), a noninvasive imaging technology, enables detailed visualization of skin layers. This study employed a deep learning approach using the U-Net architecture to analyze OCT images, specifically focusing on the stratum corneum (SC) and epidermis (ED) layers. Utilizing a dataset of 10,000 images, the U-Net model achieved 92% accuracy in predicting SC and 95% accuracy for ED. These results highlight the potential of this method for accurate 3D visualization of SC and ED layers, paving the way for precise assessment of skin layer thickness and its changes in medical applications.

Original languageEnglish (US)
Title of host publication2025 IEEE International Conference on Consumer Electronics, ICCE 2025
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331521165
DOIs
StatePublished - 2025
Event2025 IEEE International Conference on Consumer Electronics, ICCE 2025 - Las Vegas, United States
Duration: Jan 11 2025Jan 14 2025

Publication series

NameDigest of Technical Papers - IEEE International Conference on Consumer Electronics
ISSN (Print)0747-668X
ISSN (Electronic)2159-1423

Conference

Conference2025 IEEE International Conference on Consumer Electronics, ICCE 2025
Country/TerritoryUnited States
CityLas Vegas
Period1/11/251/14/25

Keywords

  • Deep Learning
  • Epidermis
  • Prediction
  • Stratum Corneum
  • U-Net

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Artificial Intelligence Based Segmentation for Skin-Layer Visualization from Optical Coherence Tomography Images'. Together they form a unique fingerprint.

Cite this