Department of Radiology, Seoul National University Hospital, 101 Daehak-ro, Jongno-gu, Seoul, 03080, South Korea.
Department of Radiology, Seoul National University College of Medicine, 101 Daehak-ro, Jongno-gu, Seoul, 03080, South Korea.
Eur Radiol. 2024 May;34(5):3431-3443. doi: 10.1007/s00330-023-10306-x. Epub 2023 Oct 20.
To develop and validate CT-based deep learning (DL) models that learn morphological and histopathological features for lung adenocarcinoma prognostication, and to compare them with a previously developed DL discrete-time survival model.
DL models were trained to simultaneously predict five morphological and histopathological features using preoperative chest CT scans from patients with resected lung adenocarcinomas. The DL score was validated in temporal and external test sets, with freedom from recurrence (FFR) and overall survival (OS) as outcomes. Discrimination was evaluated using the time-dependent area under the receiver operating characteristic curve (TD-AUC) and compared with the DL discrete-time survival model. Additionally, we performed multivariable Cox regression analysis.
In the temporal test set (640 patients; median age, 64 years), the TD-AUC was 0.79 for 5-year FFR and 0.73 for 5-year OS. In the external test set (846 patients; median age, 65 years), the TD-AUC was 0.71 for 5-year OS, equivalent to the pathologic stage (0.71 vs. 0.71 [p = 0.74]). The prognostic value of the DL score was independent of clinical factors (adjusted per-percentage hazard ratio for FFR (temporal test), 1.02 [95% CI: 1.01-1.03; p < 0.001]; OS (temporal test), 1.01 [95% CI: 1.002-1.02; p = 0.01]; OS (external test), 1.01 [95% CI: 1.005-1.02; p < 0.001]). Our model showed a higher TD-AUC than the DL discrete-time survival model, but without statistical significance (2.5-year OS: 0.73 vs. 0.68; p = 0.13).
The CT-based prognostic score from collective deep learning of morphological and histopathological features showed potential in predicting survival in lung adenocarcinomas.
Collective CT-based deep learning of morphological and histopathological features presents potential for enhancing lung adenocarcinoma prognostication and optimizing pre-/postoperative management.
• A CT-based prognostic model was developed using collective deep learning of morphological and histopathological features from preoperative CT scans of 3181 patients with resected lung adenocarcinoma. • The prognostic performance of the model was comparable-to-higher performance than the pathologic T category or stage. • Our approach yielded a higher discrimination performance than the direct survival prediction model, but without statistical significance (0.73 vs. 0.68; p=0.13).
开发并验证基于 CT 的深度学习 (DL) 模型，用于预测肺腺癌的形态学和组织病理学特征，并将其与之前开发的 DL 离散时间生存模型进行比较。
使用接受肺腺癌切除术患者的术前胸部 CT 扫描，训练 DL 模型以同时预测 5 种形态学和组织病理学特征。DL 评分在时间和外部测试集中进行验证，以无复发生存率 (FFR) 和总生存率 (OS) 为结局。使用时间依赖性接收器工作特征曲线下面积 (TD-AUC) 评估判别能力，并与 DL 离散时间生存模型进行比较。此外，我们还进行了多变量 Cox 回归分析。
在时间测试集（640 例患者；中位年龄 64 岁）中，5 年 FFR 的 TD-AUC 为 0.79，5 年 OS 的 TD-AUC 为 0.73。在外部测试集（846 例患者；中位年龄 65 岁）中，5 年 OS 的 TD-AUC 为 0.71，与病理分期相当（0.71 比 0.71 [p=0.74]）。DL 评分的预后价值独立于临床因素（FFR（时间测试）的调整后每百分比风险比，1.02 [95%CI：1.01-1.03；p<0.001]；OS（时间测试），1.01 [95%CI：1.002-1.02；p=0.01]；OS（外部测试），1.01 [95%CI：1.005-1.02；p<0.001]）。我们的模型显示出比 DL 离散时间生存模型更高的 TD-AUC，但无统计学意义（2.5 年 OS：0.73 比 0.68；p=0.13）。
基于 CT 的形态学和组织病理学特征的集体深度学习预测评分在预测肺腺癌生存方面具有潜力。
基于 CT 的形态学和组织病理学特征的集体深度学习为增强肺腺癌的预后预测和优化术前/术后管理提供了潜力。
基于术前 CT 扫描的形态学和组织病理学特征，使用集体深度学习方法，开发了一种 CT 预测模型，该模型用于预测 3181 例接受肺腺癌切除术患者的生存情况。
该模型的预测性能与病理 T 分期或分期相当或优于其预测性能。
与直接生存预测模型相比，我们的方法具有更高的判别性能，但无统计学意义（0.73 比 0.68；p=0.13）。
