State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Sun Yat-sen University Cancer Center, Guangzhou 510060, China; Department of Information, Sun Yat-Sen University Cancer Center, Guangzhou 510060, China.
State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Sun Yat-sen University Cancer Center, Guangzhou 510060, China; Department of Information, Sun Yat-Sen University Cancer Center, Guangzhou 510060, China; Precision Medicine Center, Sun Yat-Sen University, Guangzhou 510060, China.
Comput Methods Programs Biomed. 2022 Apr;217:106702. doi: 10.1016/j.cmpb.2022.106702. Epub 2022 Feb 16.
Administration of contrast is not desirable for all cases in clinical setting, and no consensus in sequence selection for deep learning model development has been achieved, thus we aim to explore whether contrast-enhanced magnetic resonance imaging (ceMRI) can be substituted in the identification and segmentation of nasopharyngeal carcinoma (NPC) with the aid of deep learning models in a large-scale cohort.
A total of 4478 eligible individuals were randomly split into training, validation and test sets, and self-constrained 3D DenseNet and V-Net models were developed using axial T1-weighted imaging (T1WI), T2WI or enhanced T1WI (T1WIC) images separately. The differential diagnostic performance between NPC and benign hyperplasia were compared among models using chi-square test. Segmentation evaluation metrics, including dice similarity coefficient (DSC) and average surface distance (ASD), were compared using paired student's t-test between T1WIC and T1WI or T2WI models or M_T1/T2, a merged output of malignant region derived from T1WI and T2WI models.
All models exhibited similar satisfactory diagnostic performance in discriminating NPC from benign hyperplasia, all attaining overall accuracy over 99.00% in all T stages of NPC. And T1WIC model exhibited similar average DSC and ASD with those of M_T1/T2 (DSC, 0.768±0.070 vs 0.764±0.070; ASD, 1.573±10.954 mm vs 1.626±10.975 mm 1.626±0.975 mm vs 1.573±0.954 mm, all p > 0.0167) in primary NPC using DenseNet, but yielded a significantly higher DSC and lower ASD than either T1WI model or T2WI model (DSC, 0.759±0.065 or 0.755±0.071; ASD, 1.661±0.898 mm or 1.722±1.133 mm, respectively, all p < 0.01) in the entire test set of NPC cohort. Moreover, the average DSCs and ASDs were not statistically significant between T1WIC model and M_T1/T2 in both.
在临床环境中，并非所有情况下都需要使用造影剂，而且对于深度学习模型开发的序列选择也没有达成共识，因此，我们旨在探索在大型队列中，借助深度学习模型，是否可以在不使用对比增强磁共振成像（ceMRI）的情况下识别和分割鼻咽癌（NPC）。
将 4478 名符合条件的个体随机分为训练集、验证集和测试集，并分别使用轴向 T1 加权成像（T1WI）、T2WI 或增强 T1WI（T1WIC）图像开发自约束 3D DenseNet 和 V-Net 模型。使用卡方检验比较模型在 NPC 与良性增生之间的鉴别诊断性能。使用配对学生 t 检验比较 T1WIC 与 T1WI 或 T2WI 模型或 M_T1/T2（源自 T1WI 和 T2WI 模型的恶性区域的合并输出）之间的分割评估指标，包括骰子相似系数（DSC）和平均表面距离（ASD）。
所有模型在鉴别 NPC 与良性增生方面均表现出相似的令人满意的诊断性能，在所有 NPC 的 T 分期中均达到了总体准确率均超过 99.00%。在原发性 NPC 中，使用 DenseNet 时，T1WIC 模型的平均 DSC 和 ASD 与 M_T1/T2 相似（DSC：0.768±0.070 比 0.764±0.070；ASD：1.573±10.954 mm 比 1.626±10.975 mm；1.626±0.975 mm 比 1.573±0.954 mm，均 p>0.0167），但与 T1WI 或 T2WI 模型相比，其 DSC 更高，ASD 更低（DSC：0.759±0.065 或 0.755±0.071；ASD：1.661±0.898 mm 或 1.722±1.133 mm，均 p<0.01）。此外，在整个 NPC 队列的测试集中，T1WIC 模型与 M_T1/T2 之间的平均 DSCs 和 ASDs 均无统计学差异。
