Cui Sunan, Luo Yi, Hsin Tseng Huan, Ten Haken Randall K, El Naqa Issam
Applied Physics Program, University of Michigan, Ann Arbor, MI 48109, USA,
Department of Radiation Oncology, University of Michigan, Medical School, MI 48103, USA,
IEEE Trans Radiat Plasma Med Sci. 2019 Mar;3(2):242-249. doi: 10.1109/TRPMS.2018.2884134. Epub 2018 Nov 29.
In this study, we investigated the application of artificial neural networks (ANNs) with composite architectures into the prediction of local control (LC) of lung cancer patients after radiotherapy. The motivation of this study was to take advantage of the temporal associations among longitudinal (sequential) data to improve the predictive performance of outcome models under the circumstance of limited sample sizes. Two composite architectures: (1) a one dimension (1D) convolutional + fully connected and (2) a locally-connected+ fully connected architectures were implemented for this purpose. Compared with the fully-connected architecture (multi-layer perceptron [MLP]), our composite architectures yielded better predictive performance of LC in lung cancer patients who received radiotherapy. Specifically, in a cohort of 98 patients (29 patients failed locally), the composite architecture of 1D convolutional layers and fully-connected layers achieved an AUC (area under receiver operating characteristic curve) of 0.83 (95% confidence interval (CI): 0.8070.841) with 18 features (14 features are longitudinal data). Whereas, the composite architecture of locally- connected layers and fully-connected layers achieved an AUC of 0.80 (95%CI: 0.7750.811). Both outperformed an MLP in the prediction performance with the same set of features, which achieved an AUC of 0.78 (95%CI: 0.751~0.790); (P-values for differences in AUC using the DeLong tests were 1.609 × 10and 1.407 × 10, respectively).
在本研究中，我们调查了具有复合架构的人工神经网络（ANN）在预测肺癌患者放疗后局部控制（LC）方面的应用。本研究的动机是利用纵向（序列）数据之间的时间关联，以在样本量有限的情况下提高结局模型的预测性能。为此实现了两种复合架构：（1）一维（1D）卷积+全连接架构，以及（2）局部连接+全连接架构。与全连接架构（多层感知器[MLP]）相比，我们的复合架构在接受放疗的肺癌患者中对LC产生了更好的预测性能。具体而言，在一个包含98名患者的队列中（29名患者局部失败），1D卷积层和全连接层的复合架构利用18个特征（14个特征为纵向数据）实现了0.83的受试者操作特征曲线下面积（AUC）（95%置信区间[CI]：0.8070.841）。而局部连接层和全连接层的复合架构实现了0.80的AUC（95%CI：0.7750.811）。在使用相同特征集的预测性能方面，这两种架构均优于MLP，后者实现了0.78的AUC（95%CI：0.751~0.790）；（使用DeLong检验的AUC差异的P值分别为1.609×10和1.407×10）。
