Related Experiment Video
Updated: Jul 15, 2025
06:21
Diffuse Optical Spectroscopy for the Quantitative Assessment of Acute Ionizing Radiation Induced Skin Toxicity Using a Mouse Model
Published on: May 27, 2016
8.2K
Development and performance validation of a low-cost algorithms-based hyperspectral imaging system for
Shicheng Hao1, Ying Xiong1, Sisi Guo1
1Key Laboratory of Photoelectronic Imaging Technology and System of Ministry of Education of China, School of Optics and Photonics, Beijing Institute of Technology, Beijing 100081, China.
Biomedical Optics Express
|October 4, 2023
Summary
A new low-cost hyperspectral imaging system accurately assesses radiodermatitis, a common side effect of radiotherapy. This system uses algorithms and Monte Carlo simulations to provide objective grading for improved patient management.
Area of Science:
- Medical Imaging
- Biophotonics
- Computational Biology
Background:
- Radiotherapy (RT) is a cornerstone of cancer treatment.
- Radiodermatitis, a severe skin reaction affecting 95% of patients, necessitates accurate grading for timely management.
- Current radiodermatitis assessment tools lack objectivity and reliability.
Purpose of the Study:
- To develop and validate a low-cost algorithms-based hyperspectral imaging (aHSI) system for radiodermatitis assessment.
- To enable objective and reliable grading of radiodermatitis.
- To explore the potential of aHSI in quantifying physiological parameters related to skin changes.
Main Methods:
- Development of a low-cost multispectral imaging (MSI) system using LEDs and a CMOS camera.
- Acquisition of multi-spectra and derivation of algorithms-based hyper-spectra (1 nm resolution) via Monte Carlo (MC) simulations.
- Measurement of artificially induced erythema in volunteers and derivation of skin physiological parameters (BVF, oxygen saturation) using MC simulations.
- Classification of radiodermatitis using a 1D-convolution neural network (CNN) on derived hyper-spectra.
Main Results:
- The aHSI system achieved a spectral resolution of 1 nm.
- MC simulations successfully derived blood volume fraction (BVF) and oxygen saturation, showing significant increases with erythema (P < 0.001).
- The 1D-CNN model achieved an overall classification accuracy of 93.1% for radiodermatitis grading.
Conclusions:
- The developed low-cost aHSI system demonstrates high accuracy in assessing radiodermatitis.
- The system shows significant potential for objective and reliable radiodermatitis grading, aiding clinical management.
- This technology offers a cost-effective solution for improving cancer patient care by monitoring RT side effects.
