The 12‑lead ECG provides an excellent substrate for artificial intelligence (AI) enabled prediction of various cardiovascular diseases. However, a measure of prediction certainty is lacking. To assess a novel approach for estimating certainty of AI-ECG predictions. Two convolutional neural networks (CNN) were developed to predict patient age and sex. Model 1 applied a 5 s sliding time-window, allowing multiple CNN predictions. The consistency of the output values, expressed as interquartile range (IQR), was used to estimate prediction certainty. Model 2 was trained on the full 10s ECG signal, resulting in a single CNN point prediction value. Performance was evaluated on an internal test set and externally validated on the PTB-XL dataset. Both CNNs were trained on 269,979 standard 12‑lead ECGs (82,477 patients). Model 1 showed higher accuracy for both age and sex prediction (mean absolute error, MAE 6.9 ± 6.3 years vs. 7.7 ± 6.3 years and AUC 0.946 vs. 0.916, respectively, < 0.001 for both). The IQR of multiple CNN output values allowed to differentiate between high and low accuracy of ECG based predictions ( < 0.001 for both). Among 10% of patients with narrowest IQR, sex prediction accuracy increased from 65.4% to 99.2%, and MAE of age prediction decreased from 9.7 to 4.1 years compared to the 10% with widest IQR. Accuracy and estimation of prediction certainty of model 1 remained true in the external validation dataset. Sliding window-based approach improves ECG based prediction of age and sex and may aid in addressing the challenge of prediction certainty estimation.

中文翻译：

---

人工智能心电图的确定性预测

12 导联心电图为人工智能 (AI) 预测各种心血管疾病提供了极好的基础。然而，缺乏预测确定性的衡量标准。评估一种估计 AI-ECG 预测确定性的新方法。开发了两个卷积神经网络（CNN）来预测患者的年龄和性别。模型 1 应用了 5 秒的滑动时间窗口，允许多个 CNN 预测。输出值的一致性（以四分位距 (IQR) 表示）用于估计预测的确定性。模型 2 在完整的 10 秒 ECG 信号上进行训练，产生单个 CNN 点预测值。性能在内部测试集上进行评估，并在 PTB-XL 数据集上进行外部验证。两个 CNN 均接受了 269,979 个标准 12 导联心电图（82,477 名患者）的训练。模型 1 在年龄和性别预测方面显示出更高的准确性（平均绝对误差，MAE 分别为 6.9 ± 6.3 岁 vs. 7.7 ± 6.3 岁，AUC 分别为 0.946 vs. 0.916，两者均 < 0.001）。多个 CNN 输出值的 IQR 允许区分基于 ECG 的预测的高准确度和低准确度（两者均 < 0.001）。在 IQR 最窄的 10% 患者中，与 IQR 最宽的 10% 患者相比，性别预测准确率从 65.4% 提高到 99.2%，年龄预测的 MAE 从 9.7 岁下降到 4.1 岁。模型 1 的预测确定性的准确性和估计在外部验证数据集中保持正确。基于滑动窗口的方法改进了基于心电图的年龄和性别预测，并可能有助于解决预测确定性估计的挑战。