Park Sehoon, Chung Soomin, Kim Yisak, Yang Sun-Ah, Kwon Soie, Cho Jeong Min, Lee Min Jae, Cho Eunbyeol, Ryu Jiwon, Kim Sejoong, Lee Jeonghwan, Yoon Hyung Jin, Choi Edward, Kim Kwangsoo, Lee Hajeong
Department of Internal Medicine, Seoul National University Hospital, Seoul, Korea.
Interdisciplinary Program in Bioengineering, Seoul National University, Seoul, Korea.
PLoS Med. 2025 Apr 29;22(4):e1004566. doi: 10.1371/journal.pmed.1004566. eCollection 2025 Apr.
Postoperative acute kidney injury (PO-AKI) prediction models for non-cardiac major surgeries typically rely solely on preoperative clinical characteristics.
In this study, we developed and externally validated a deep-learning-based model that integrates preoperative data with minute-scale intraoperative vital signs to predict PO-AKI. Using data from three hospitals, we constructed a convolutional neural network-based EfficientNet framework to analyze intraoperative data and created an ensemble model incorporating 103 baseline variables of demographics, medication use, comorbidities, and surgery-related characteristics. Model performance was compared with the conventional SPARK model from a previous study. Among 110,696 patients, 51,345 were included in the development cohort, and 59,351 in the external validation cohorts. The median age of the cohorts was 60, 61, and 66 years, respectively, with males comprising 54.9%, 50.8%, and 42.7% of each cohort. The intraoperative vital sign-based model demonstrated comparable predictive power (AUROC (Area Under the Receiver Operating Characteristic Curve): discovery cohort 0.707, validation cohort 0.637 and 0.607) to preoperative-only models (AUROC: discovery cohort 0.724, validation cohort 0.697 and 0.745). Adding 11 key clinical variables (e.g., age, sex, estimated glomerular filtration rate (eGFR), albuminuria, hyponatremia, hypoalbuminemia, anemia, diabetes, renin-angiotensin-aldosterone inhibitors, emergency surgery, and the estimated surgery time) improved the model's performance (AUROC: discovery cohort 0.765, validation cohort 0.716 and 0.761). The ensembled deep-learning model integrating both preoperative and intraoperative data achieved the highest predictive accuracy (AUROC: discovery cohort 0.795, validation cohort 0.762 and 0.786), outperforming the conventional SPARK model. The retrospective design in a single-nation cohort with non-inclusion of some potential AKI-associated variables is the main limitation of this study.
This deep-learning-based PO-AKI risk prediction model provides a comprehensive approach to evaluating PO-AKI risk prediction by combining preoperative clinical data with real-time intraoperative vital sign information, offering enhanced predictive performance for better clinical decision-making.
非心脏大手术的术后急性肾损伤（PO-AKI）预测模型通常仅依赖术前临床特征。
在本研究中，我们开发并外部验证了一种基于深度学习的模型，该模型将术前数据与分钟级术中生命体征相结合以预测PO-AKI。利用来自三家医院的数据，我们构建了基于卷积神经网络的EfficientNet框架来分析术中数据，并创建了一个整合了人口统计学、用药情况、合并症和手术相关特征等103个基线变量的集成模型。将模型性能与先前研究中的传统SPARK模型进行比较。在110,696例患者中，51,345例纳入开发队列，59,351例纳入外部验证队列。各队列的中位年龄分别为60岁、61岁和66岁，男性分别占各队列的54.9%、50.8%和42.7%。基于术中生命体征的模型显示出与仅基于术前的模型相当的预测能力（受试者操作特征曲线下面积（AUROC）：发现队列0.707，验证队列0.637和0.607）（AUROC：发现队列0.724，验证队列0.697和0.745）。添加11个关键临床变量（如年龄、性别、估计肾小球滤过率（eGFR）、蛋白尿、低钠血症、低白蛋白血症、贫血、糖尿病、肾素-血管紧张素-醛固酮抑制剂、急诊手术和估计手术时间）可改善模型性能（AUROC：发现队列0.765，验证队列0.716和0.761）。整合术前和术中数据的集成深度学习模型实现了最高的预测准确性（AUROC：发现队列0.795，验证队列0.762和0.786），优于传统的SPARK模型。本研究的主要局限性在于单国家队列的回顾性设计以及未纳入一些潜在的与急性肾损伤相关的变量。
这种基于深度学习的PO-AKI风险预测模型通过将术前临床数据与实时术中生命体征信息相结合，为评估PO-AKI风险预测提供了一种全面的方法，具有增强的预测性能，有助于更好的临床决策。
