Related Experiment Video
Updated: Aug 9, 2025
06:38
Induction of Ocular Surface Inflammation and Collection of Involved Tissues
Published on: August 4, 2022
2.3K
Quantifying Inflammatory Response and Drug-Aided Resolution in an Atopic Dermatitis Model with Deep Learning.
Daniel A Greenfield1, Amin Feizpour2, Conor L Evans1
1Wellman Center for Photomedicine, Harvard Medical School/Massachusetts General Hospital, Boston, Massachusetts, USA; Biophysics PhD Program, Harvard University, The Graduate School of Arts and Sciences, Boston, Massachusetts, USA.
The Journal of Investigative Dermatology
|February 21, 2023
Summary
This study introduces a noninvasive method using deep learning and advanced imaging to quantify skin inflammation in atopic dermatitis. This approach offers a cellular-level understanding, improving disease assessment for better treatment development.
Area of Science:
- Dermatology
- Medical Imaging
- Computational Biology
Background:
- Accurate quantification of atopic dermatitis (AD) inflammation is challenging due to limitations of macroscale observations.
- Current diagnostic methods, like biopsies, are invasive and do not fully capture cellular-level disease processes.
- There is a need for noninvasive techniques to assess skin inflammation for improved AD diagnosis and treatment.
Purpose of the Study:
- To develop and validate a noninvasive, image-based method for quantifying skin inflammation in an atopic dermatitis mouse model.
- To utilize deep learning analysis of advanced imaging techniques for cellular-level insights into AD.
- To establish a quantitative workflow for assessing disease onset and resolution in AD.
Main Methods:
- Employed coherent anti-Stokes Raman scattering (CARS) and stimulated Raman scattering (SRS) imaging for high-resolution visualization.
- Applied deep learning algorithms for cellular-level analysis of imaging data.
- Developed a quantification method based on morphological and physiological measurements to generate timepoint-specific disease scores.
Main Results:
- Successfully achieved noninvasive quantification of inflammation in a mouse model of atopic dermatitis.
- Demonstrated the ability to generate timepoint-specific disease scores using the developed image-based deep learning approach.
- Provided cellular-level insights into the inflammatory processes of atopic dermatitis.
Conclusions:
- The developed noninvasive quantification method provides a cellular-level assessment of skin inflammation in atopic dermatitis.
- This approach overcomes the limitations of traditional invasive methods, enabling broader clinical study enrollment.
- The workflow is poised for application in future clinical studies, advancing the diagnosis and treatment of atopic dermatitis.
