AI hybrid survival assessment for advanced heart failure patients with renal dysfunction

• Published in: Nature communications Vol. 15; no. 1; pp. 6756 - 21
• Main Authors: Zhang, Ge, Wang, Zeyu, Tong, Zhuang, Qin, Zhen, Su, Chang, Li, Demin, Xu, Shuai, Li, Kaixiang, Zhou, Zhaokai, Xu, Yudi, Zhang, Shiqian, Wu, Ruhao, Li, Teng, Zheng, Youyang, Zhang, Jinying, Cheng, Ke, Tang, Junnan
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 08.08.2024; Nature Publishing Group; Nature Portfolio
• Subjects: 631/1647/794; 692/4019; 692/499; 692/53/2422; 692/699/75/230; Aged; Artificial Intelligence; Biomarkers; Cardiovascular diseases; Congestive heart failure; Empowerment; Female; Health care; Health care facilities; Heart diseases; Heart failure; Heart Failure - mortality; Heart Failure - physiopathology; Humanities and Social Sciences; Humans; Male; Medical personnel; Medical prognosis; Middle Aged; Modelling; Monitoring; multidisciplinary; Patients; Prognosis; Renal failure; Renal function; Renal Insufficiency - diagnosis; Renal Insufficiency - mortality; Renal Insufficiency - physiopathology; Research facilities; Risk assessment; Risk Assessment - methods; Science; Science (multidisciplinary); Survival; Survival Analysis; Telemedicine
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-024-50415-9

Renal dysfunction (RD) often characterizes the worse course of patients with advanced heart failure (AHF). Many prognosis assessments are hindered by researcher biases, redundant predictors, and lack of clinical applicability. In this study, we enroll 1736 AHF/RD patients, including data from Henan Province Clinical Research Center for Cardiovascular Diseases (which encompasses 11 hospital subcenters), and Beth Israel Deaconess Medical Center. We developed an AI hybrid modeling framework, assembling 12 learners with different feature selection paradigms to expand modeling schemes. The optimized strategy is identified from 132 potential schemes to establish an explainable survival assessment system: AIHFLevel. The conditional inference survival tree determines a probability threshold for prognostic stratification. The evaluation confirmed the system’s robustness in discrimination, calibration, generalization, and clinical implications. AIHFLevel outperforms existing models, clinical features, and biomarkers. We also launch an open and user-friendly website www.hf-ai-survival.com , empowering healthcare professionals with enhanced tools for continuous risk monitoring and precise risk profiling. Here the authors show an AI-powered assessment system (AIHFLevel, www.hf-ai-survival.com ) empowering healthcare professionals for continuous risk monitoring and prognosis assessment of patients who have advanced heart failure with renal dysfunction.