Room: ePoster Forums
Purpose: To visualize changes in blood-flow between non-irradiated and irradiated skin tissue through the development of a user-friendly software framework for the clinical research utilization of Eulerian video amplification in radiation oncology.
Methods: MATLAB code was written to implement a technique developed by the M.I.T. imaging lab to amplify colors present in video sequences. This technique, Eulerian amplification, theoretically allows the visualization of skin color changes as a result of the inflow and outflow of blood during cardiac cycles. The developed software established a means to segment non-irradiated tissue from irradiated tissue in the video frame and to adjust temporal parameters of the amplification. In a pilot test of the software, high-resolution videos of irradiated mice lasting less that than 5 seconds were acquired and both irradiated and surrounding skin were amplified using Eulerian techniques. After applying Eulerian amplification to the dynamic videos of mice throughout their course of radiotherapy for GBM tumors, we extracted quantitative information regarding blood flow and color patterns.
Results: The software provided a platform to successfully analyze periodic non-visible color variations in skin corresponding to blood flow based on the pilot mouse study. The software allowed user-selected points of interest to be tracked through the image time series to extract cardiac frequencies and exclude breathing frequencies to determine the optimal window for amplification. From pulse color variations that were indiscernible in the original video but notable in the amplified video, we were able to extract metrics comparing pixel value variations in irradiated and non-irradiated regions.
Conclusion: User-friendly software was created to provide a technique to assess skin changes during radiotherapy that will be useful in the clinic for detecting erythema before clinical manifestation in breast cancer patients who are prone to severe skin toxicities due to complex tissue variations and inhomogeneities in dose distribution.