Avanzo Michele, Pirrone Giovanni, Vinante Lorenzo, Caroli Angela, Stancanello Joseph, Drigo Annalisa, Massarut Samuele, Mileto Mario, Urbani Martina, Trovo Marco, El Naqa Issam, De Paoli Antonino, Sartor Giovanna
Department of Medical Physics, Centro di Riferimento Oncologico di Aviano (CRO) IRCCS, Aviano, Italy.
Department of Radiation Oncology, Centro di Riferimento Oncologico di Aviano (CRO) IRCCS, Aviano, Italy.
Front Oncol. 2020 Apr 21;10:490. doi: 10.3389/fonc.2020.00490. eCollection 2020.
to predict the occurrence of late subcutaneous radiation induced fibrosis (RIF) after partial breast irradiation (PBI) for breast carcinoma by using machine learning (ML) models and radiomic features from 3D Biologically Effective Dose (3D-BED) and Relative Electron Density (3D-RED). 165 patients underwent external PBI following a hypo-fractionation protocol consisting of 40 Gy/10 fractions, 35 Gy/7 fractions, and 28 Gy/4 fractions, for 73, 60, and 32 patients, respectively. Physicians evaluated toxicity at regular intervals by the Common Terminology Adverse Events (CTAE) version 4.0. RIF was assessed every 3 months after the completion of radiation course and scored prospectively. RIF was experienced by 41 (24.8%) patients after average 5 years of follow up. The Hounsfield Units (HU) of the CT-images were converted into relative electron density (3D-RED) and Dose maps into Biologically Effective Dose (3D-BED), respectively. Shape, first-order and textural features of 3D-RED and 3D-BED were calculated in the planning target volume (PTV) and breast. Clinical and demographic variables were also considered (954 features in total). Imbalance of the dataset was addressed by data augmentation using ADASYN technique. A subset of non-redundant features that best predict the data was identified by sequential feature selection. Support Vector Machines (SVM), ensemble machine learning (EML) using various aggregation algorithms and Naive Bayes (NB) classifiers were trained on patient dataset to predict RIF occurrence. Models were assessed using sensitivity and specificity of the ML classifiers and the area under the receiver operator characteristic curve (AUC) of the score functions in repeated 5-fold cross validation on the augmented dataset. The SVM model with seven features was preferred for RIF prediction and scored sensitivity 0.83 (95% CI 0.80-0.86), specificity 0.75 (95% CI 0.71-0.77) and AUC of the score function 0.86 (0.85-0.88) on cross-validation. The selected features included cluster shade and Run Length Non-uniformity of breast 3D-BED, kurtosis and cluster shade from PTV 3D-RED, and 10th percentile of PTV 3D-BED. Textures extracted from 3D-BED and 3D-RED in the breast and PTV can predict late RIF and may help better select patient candidates to exclusive PBI.
利用机器学习（ML）模型以及来自三维生物等效剂量（3D-BED）和相对电子密度（3D-RED）的影像组学特征，预测乳腺癌部分乳腺照射（PBI）后迟发性皮下放射性纤维化（RIF）的发生情况。165例患者接受了大分割方案的外照射PBI，其中73例、60例和32例患者分别接受了40 Gy/10次、35 Gy/7次和28 Gy/4次的照射。医生按照《常见不良反应事件术语标准》（CTAE）第4.0版定期评估毒性。放疗疗程结束后每3个月评估一次RIF，并进行前瞻性评分。平均随访5年后，41例（24.8%）患者出现RIF。CT图像的亨氏单位（HU）分别转换为相对电子密度（3D-RED），剂量图转换为生物等效剂量（3D-BED）。在计划靶区（PTV）和乳腺中计算3D-RED和3D-BED的形状、一阶和纹理特征。还考虑了临床和人口统计学变量（共954个特征）。采用ADASYN技术进行数据增强，解决数据集不平衡问题。通过序列特征选择，识别出最能预测数据的非冗余特征子集。在患者数据集上训练支持向量机（SVM）、使用各种聚合算法的集成机器学习（EML）和朴素贝叶斯（NB）分类器，以预测RIF的发生。在增强数据集上进行重复5折交叉验证时，使用ML分类器的敏感性和特异性以及评分函数的受试者操作特征曲线下面积（AUC）评估模型。具有七个特征的SVM模型在RIF预测中表现最佳，在交叉验证中的敏感性为0.83（95%CI 0.80-0.86），特异性为0.75（95%CI 0.71-0.77），评分函数的AUC为0.86（0.85-0.88）。所选特征包括乳腺3D-BED的聚类阴影和游程不均匀性、PTV 3D-RED的峰度和聚类阴影，以及PTV 3D-BED的第10百分位数。从乳腺和PTV的3D-BED和3D-RED中提取的纹理可以预测迟发性RIF，并可能有助于更好地选择适合单纯PBI的患者。
