Stratum corneum nanotexture feature detection using deep learning and spatial analysis: a non-invasive tool for skin barrier assessment
This repository presents an objective, quantifiable method for assessing atopic dermatitis (AD) severity. The program integrates deep learning object detection with spatial analysis algorithms to accurately calculate the density of circular nano-size objects (CNOs), termed the Effective Corneocyte Topographical Index (ECTI). The ECTI demonstrates remarkable robustness in overcoming the inherent challenges of nano-imaging, such as environmental noise and structural occlusions on the corneocyte surface, further enhancing its applicability in clinical settings.
Dependencies
- Python 3.9+
- matplotlib
- numpy
- opencv-python
- scipy
- scikit-image
- ultralytics
- scikit-learn
- customtkinter
Directories
AD_Assessment_GUI.zip
contains a cross-platform executable GUI, sample data, and a tutorial video.utils/Img_Preprocessing.py
demonstrates the image enhancement algorithms applied to the corneocyte nanotexture images.
Usage
-
Execution via cross-platform executable GUI
- Download AD_Assessment_GUI.zip
- Run
AD_Assessment_GUI.exe
- Analysis results will be saved within the selected path in a folder titled
CNO_Detection
-
Execution via python script
- Install packages in terminal:
pip install -r requirements.txt
- Run
AD_Assessment_GUI.py
- Analysis results will be saved within the selected path in a folder titled
CNO_Detection
- Install packages in terminal:
Executable
-
Install PyInstaller in terminal:
pip install pyinstaller
-
Run command in terminal:
pyinstaller --onedir .\AD_Assessment_GUI.py
Performance
Model | Test Size | #Parameter (M) | FLOPs (G) | AP50 (%) | AP50-95 (%) | Latency (ms) |
---|---|---|---|---|---|---|
YOLOv10-N | 512 | 2.7 | 8.2 | 89.6 | 51.4 | 3.3 |
YOLOv10-S | 512 | 8.0 | 24.4 | 90.8 | 55.5 | 4.58 |
YOLOv10-M | 512 | 16.5 | 63.4 | 91.3 | 59.7 | 7.17 |
YOLOv10-B | 512 | 20.4 | 97.7 | 91.1 | 62.5 | 7.58 |
YOLOv10-L | 512 | 25.7 | 126.3 | 91.4 | 63.2 | 9.01 |
YOLOv10-X | 512 | 31.6 | 169.8 | 91.2 | 62.9 | 10.95 |
RT-DETRv2-S | 512 | 20.0 | 60.0 | 87.6 | 39.6 | 5.51 |
RT-DETRv2-M | 512 | 31.0 | 100.0 | 84.0 | 37.2 | 7.48 |
RT-DETRv2-L | 512 | 42.0 | 136.0 | 84.3 | 33.4 | 13.50 |
RT-DETRv2-X | 512 | 76.0 | 259.0 | 83.3 | 32.0 | 21.15 |
Dataset
The corneocyte nanotexture dataset is available for download at the following link: Corneocyte Nanotexture Dataset.
Contributions
[1] Liao, H-S., Wang, J-H., Raun, E., Nørgaard, L. O., Dons, F. E., & Hwu, E. E-T. (2022). Atopic Dermatitis Severity Assessment using High-Speed Dermal Atomic Force Microscope. Abstract from AFM BioMed Conference 2022, Nagoya-Okazaki, Japan.
[2] Pereda, J., Liao, H-S., Werner, C., Wang, J-H., Huang, K-Y., Raun, E., Nørgaard, L. O., Dons, F. E., & Hwu, E. E. T. (2022). Hacking Consumer Electronics for Biomedical Imaging. Abstract from 5th Global Conference on Biomedical Engineering & Annual Meeting of TSBME, Taipei, Taiwan, Province of China.
[3] Liao, H. S., Akhtar, I., Werner, C., Slipets, R., Pereda, J., Wang, J. H., Raun, E., Nørgaard, L. O., Dons, F. E., & Hwu, E. E. T. (2022). Open-source controller for low-cost and high-speed atomic force microscopy imaging of skin corneocyte nanotextures. HardwareX, 12, [e00341]. https://doi.org/10.1016/j.ohx.2022.e00341
Contact: Jen-Hung Wang / Assoc. Professor En-Te Hwu
Version History
master @ fafd5bf (earliest) Created 11th Sep 2024 at 12:09 by Jen-Hung Wang
Update readme.md
Frozen
master
fafd5bf
Creator
Submitter
Views: 514 Downloads: 76
Created: 11th Sep 2024 at 12:09
None