Management of grade 3 acute dermatitis with moist desquamation after adjuvant chest wall radiotherapy: a case report

Quantitative evaluation of radiodermatitis following whole-breast radiotherapy with various color space models: A feasibility study

Purpose

We analyzed skin images with various color space models to objectively assess radiodermatitis severity in patients receiving whole-breast radiotherapy.

Methods

Twenty female patients diagnosed with breast cancer were enrolled prospectively. All patients received whole-breast radiotherapy without boost irradiation. Skin images for both irradiated and unirradiated breasts were recorded in red-green-blue (RGB) color space using a mobile skin analysis device. For longitudinal analysis, the images were acquired before radiotherapy (RTbefore), approximately 7 days after the first fraction of radiotherapy (RT7days), RT14days, and approximately 10 days after radiotherapy completion (RTafter). Four color space models (RGB, hue-saturation-value (HSV), L*a*b*, and YCbCr models) were employed to calculate twelve color space parameters for each skin image. Skin dose measurements for irradiated breasts were performed using nanoDot optically stimulated luminescent dosimeters on the first fraction of radiotherapy. Subsequently, acute radiation dermatitis in each patient was assessed according to the Radiation Therapy Oncology Group scoring criteria at both RT14days and RTafter. Finally, several statistical analysis methods were applied to investigate the performance of the color space parameters to objectively assess the radiodermatitis.

Results

Owing to radiation-induced skin damage, R value of RGB model as well as S and V values of the HSV model for irradiated breasts increased significantly, while those for unirradiated breasts showed smaller increases. These parameters showed the longitudinal changes in color space parameters within each group and between groups over time with statistical significance. Strong correlations of the parameters for irradiated breasts at RT7days with skin doses and those at RTafter were observed with statistical significance.

Conclusion

The R value of RGB model as well as the S and V values of HSV model showed relatively better performance in evaluating the acute radiation dermatitis. These color space parameters could therefore serve as useful tools to assess radiodermatitis severity in a dose-dependent manner.

A data science approach for early-stage prediction of Patient's susceptibility to acute side effects of advanced radiotherapy

The prediction by classification of side effects incidence in a given medical treatment is a common challenge in medical research. Machine Learning (ML) methods are widely used in the areas of risk prediction and classification. The primary objective of such algorithms is to use several features to predict dichotomous responses (e.g., disease positive/negative). Similar to statistical inference modelling, ML modelling is subject to the class imbalance problem and is affected by the majority class, increasing the false-negative rate. In this study, seventy-nine ML models were built and evaluated to classify approximately 2000 participants from 26 hospitals in eight different countries into two groups of radiotherapy (RT) side effects incidence based on recorded observations from the international study of RT related toxicity "REQUITE". We also examined the effect of sampling techniques and cost-sensitive learning methods on the models when dealing with class imbalance. The combinations of such techniques used had a significant impact on the classification. They resulted in an improvement in incidence status prediction by shifting classifiers' attention to the minority group. The best classification model for RT acute toxicity prediction was identified based on domain experts' success criteria. The Area Under Receiver Operator Characteristic curve of the models tested with an isolated dataset ranged from 0.50 to 0.77. The scale of improved results is promising and will guide further development of models to predict RT acute toxicities. One model was optimised and found to be beneficial to identify patients who are at risk of developing acute RT early-stage toxicities as a result of undergoing breast RT ensuring relevant treatment interventions can be appropriately targeted. The design of the approach presented in this paper resulted in producing a preclinical-valid prediction model. The study was developed by a multi-disciplinary collaboration of data scientists, medical physicists, oncologists and surgeons in the UK Radiotherapy Machine Learning Network.