Liang Mengdi, Liu Yuelin, Huang Yue, Ma Ge, Han Xu, Li Shuaikang, Hang Jing, Xie Hui, Chen Lin, Liu Xiaoan, Wang Shui, Xia Tiansong
Department of Breast Surgery, The First Affiliated Hospital with Nanjing Medical University, Nanjing, China.
State Key Laboratory of Ultrasound in Medicine and Engineering, College of Biomedical Engineering, Chongqing Medical University, Chongqing, China.
Int J Hyperthermia. 2025 Dec;42(1):2497824. doi: 10.1080/02656736.2025.2497824. Epub 2025 Jul 9.
To establish a predictive model for the sonication energy required for focused ultrasound surgery (FUS) of breast fibroadenomas.
This study retrospectively enrolled 87 patients with 154 benign breast tumors treated by FUS in our hospital. Radiomic analysis included 124 tumors from 69 patients, randomly split into a 3:1 ratio for training (96 cases) and validation (28 cases). Three machine learning algorithms were applied for feature selection. Then, all the selected features were used for the construction of the prediction model via four machine learning algorithms. Residual analysis and Intraclass Correlation Coefficient (ICC) analysis were performed to evaluate the performances of these four models. The importance of each feature is demonstrated by the Root Mean Square Error (RMSE) loss obtained through permutation importance measurement.
This study collected 11 clinical features and 68 ultrasound radiomics features, totaling 79 independent variables. The Bagging Tree Model, characterized by lower and stable RMSE values and high R stability with increasing features, demonstrated superior predictive accuracy and explanatory power compared to other models. At the optimal feature count, identified by the minimum RMSE, 33 features were selected for further modeling. The bagging tree model has the highest ICC value among the four models, at 0.56, with a confidence interval of (0.23, 0.77).
This study established an interpretable machine learning model that integrates clinical and ultrasound radiomics features to estimate the sonication energy in FUS treatment of breast fibroadenomas.
建立乳腺纤维瘤聚焦超声手术（FUS）所需超声能量的预测模型。
本研究回顾性纳入了我院87例接受FUS治疗的154个乳腺良性肿瘤患者。放射组学分析包括69例患者的124个肿瘤，随机按3:1比例分为训练组（96例）和验证组（28例）。应用三种机器学习算法进行特征选择。然后，通过四种机器学习算法将所有选定特征用于构建预测模型。进行残差分析和组内相关系数（ICC）分析以评估这四种模型的性能。通过排列重要性测量获得的均方根误差（RMSE）损失来证明每个特征的重要性。
本研究收集了11个临床特征和68个超声放射组学特征，共79个独立变量。与其他模型相比，Bagging树模型具有较低且稳定的RMSE值，并且随着特征增加R稳定性高，表现出卓越的预测准确性和解释力。在由最小RMSE确定的最佳特征数量下，选择33个特征进行进一步建模。Bagging树模型在四种模型中具有最高的ICC值，为0.56，置信区间为（0.23, 0.77）。
本研究建立了一个可解释的机器学习模型，该模型整合了临床和超声放射组学特征，以估计乳腺纤维瘤FUS治疗中的超声能量。
