Quantitative Radiomics Approach to Assess Acute Radiation Dermatitis in Breast Cancer Patients

Abstract

We applied a radiomics approach to skin surface images to objectively assess radiodermatitis in patients undergoing radiotherapy for breast cancer. A prospective cohort study of 20 patients was conducted. Skin surface images in normal, polarised, and UV modes were acquired using a skin analysis device before starting radiotherapy (‘before RT’), 7 days after the first treatment ('RT_D7'), on ‘RT_D14’, and 10 days after radiotherapy ended (‘after RT_D10’). Eighteen types of radiomic feature ratios were calculated. We measured skin doses in ipsilateral breasts using OSLDs on the first day of radiotherapy. Clinical evaluation of acute radiodermatitis was performed using the RTOG scoring criteria on ‘RT_D14’ and ‘after RT_D10’. Several statistical analysis methods were used to test the performance of radiomic features as indicators of radiodermatitis evaluation. As the skin was damaged by radiation, the energy for normal mode and sum variance for polarised and UV modes decreased significantly for ipsilateral breasts. The radiomic feature ratios at ‘RT_D7’ had strong correlations to skin doses and those at ‘RT_D14’ and ‘after RT_D10’ with statistical significance. The energy for normal mode and sum variance for polarised and UV modes demonstrated the potential to evaluate and predict acute radiation, which assists in its appropriate management.