School of Medicine, National Defense Medical Center, Taipei, Taiwan, ROC; School of Public Health, National Defense Medical Center, Taipei, Taiwan, ROC.
Division of Nephrology, Department of Medicine, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC.
Clin Chim Acta. 2022 Nov 1;536:126-134. doi: 10.1016/j.cca.2022.09.021. Epub 2022 Sep 24.
Abnormal serum calcium concentrations affect the heart and may alter the electrocardiogram (ECG), but the detection of hypocalcemia and hypercalcemia (collectively dyscalcemia) relies on blood laboratory tests requiring turnaround time.
The study aimed to develop a bloodless artificial intelligence (AI)-enabled (ECG) method to rapidly detect dyscalcemia and analyze its possible utility for outcome prediction.
This study collected 86,731 development, 15,611 tuning, 11,105 internal validation, and 8401 external validation ECGs from electronic medical records with at least 1 ECG associated with an albumin-adjusted calcium (aCa) value within 4 h. The main outcomes were to assess the accuracy of AI-ECG to predict aCa and follow up these patients for all-cause mortality, new-onset acute myocardial infraction (AMI), and new-onset heart failure (HF) to validate the ability of AI-ECG-aCa for previvor identification.
ECG-aCa had mean absolute errors (MAE) of 0.78/0.98 mg/dL and achieved an area under receiver operating characteristic curves (AUCs) 0.9219/0.8447 and 0.8948/0.7723 to detect severe hypercalcemia and hypocalcemia in the internal/external validation sets, respectively. Although < 20 % variance of ECG-aCa could be explained by traditional ECG features, the ECG-aCa was found to be associated with more complications. Patients with ECG-hypercalcemia but initially normal aCa were found to have a higher risk of subsequent all-cause mortality [hazard ratio (HR): 2.05, 95 % conference interval (CI): 1.55-2.70], new-onset AMI (HR: 2.88, 95 % CI: 1.72-4.83), and new-onset HF (HR: 2.02, 95 % CI: 1.38-2.97) in the internal validation set, which were also seen in external validation.
The AI-ECG-aCa may help detecting severe dyscalcemia for early diagnosis and ECG-hypercalcemia also has prognostic value for clinical outcomes (all-cause mortality and new-onset AMI and HF).
异常血清钙浓度会影响心脏，并可能改变心电图（ECG），但低钙血症和高钙血症（统称为钙异常）的检测依赖于需要周转时间的血液实验室检查。
本研究旨在开发一种无血人工智能（AI）驱动的（ECG）方法，以快速检测钙异常，并分析其对预后预测的可能效用。
本研究从电子病历中收集了 86731 份开发、15611 份调整、11105 份内部验证和 8401 份外部验证的 ECG，其中至少有 1 份 ECG 与白蛋白校正钙（aCa）值相关，其值在 4 小时内。主要结局是评估 AI-ECG 预测 aCa 的准确性，并对这些患者进行全因死亡率、新发急性心肌梗死（AMI）和新发心力衰竭（HF）的随访，以验证 AI-ECG-aCa 对预发病的识别能力。
ECG-aCa 的平均绝对误差（MAE）分别为 0.78/0.98mg/dL，在内部/外部验证集中，其获得的接受者操作特征曲线（AUC）面积分别为 0.9219/0.8447 和 0.8948/0.7723，以检测严重高钙血症和低钙血症。尽管传统 ECG 特征只能解释<20%的 ECG-aCa 方差，但发现 ECG-aCa 与更多并发症相关。在内部验证集中，ECG 高钙血症但初始 aCa 正常的患者随后发生全因死亡率的风险更高[风险比（HR）：2.05，95%置信区间（CI）：1.55-2.70]、新发 AMI（HR：2.88，95%CI：1.72-4.83）和新发 HF（HR：2.02，95%CI：1.38-2.97），在外部验证中也观察到了这一点。
AI-ECG-aCa 可能有助于早期诊断严重钙异常，ECG 高钙血症对临床结局（全因死亡率和新发 AMI 和 HF）也具有预后价值。
