人工智能在早产儿疾病诊疗中的研究进展

袁英 ,  唐翎瀚 ,  管利荣

中国当代儿科杂志 ›› 2026, Vol. 28 ›› Issue (05) : 629 -635.

PDF (550KB)
中国当代儿科杂志 ›› 2026, Vol. 28 ›› Issue (05) : 629 -635. DOI: 10.7499/j.issn.1008-8830.2507030
综述

人工智能在早产儿疾病诊疗中的研究进展

作者信息 +

Research progress in artificial intelligence for the diagnosis and management of diseases in preterm infants

Author information +
文章历史 +
PDF (563K)

摘要

人工智能(artificial intelligence, AI)技术在医学领域发展迅速,尤其在早产儿疾病诊疗中展现出重要临床价值。早产儿各器官发育不成熟,并发症发生率高,其早期预测、精准诊断和个体化治疗是临床面临的重大挑战。AI凭借强大的数据处理和模式识别能力,为早产儿疾病诊疗提供了新的解决方案,现已广泛应用于早产儿并发症预测、影像学诊断辅助、治疗方案优化及预后评估等方面,显著提高了诊疗效率和准确性。但当前AI技术在临床应用中仍存在数据质量、模型可解释性及伦理问题等局限性。该文就AI在早产儿疾病诊疗中的研究进展进行综述,探讨其应用优势、挑战及未来发展方向,为临床实践和相关研究提供参考。

Abstract

Artificial intelligence (AI) technology is developing rapidly in the medical field, particularly showing significant clinical value in the diagnosis and management of diseases in preterm infants. Preterm infants have immature organ development and a high incidence of complications; early prediction, accurate diagnosis, and individualized treatment pose major clinical challenges. With its powerful data processing and pattern recognition capabilities, AI provides new solutions for the diagnosis and management of diseases in preterm infants. It is now widely applied to the prediction of complications, imaging diagnosis, optimization of treatment plans, and prognostic evaluation for preterm infants, significantly improving diagnostic and therapeutic efficiency and accuracy. However, limitations remain in the clinical application, including data quality, model interpretability, and ethical issues. This article reviews the research progress of AI in the diagnosis and management of diseases in preterm infants, discusses its application advantages, challenges, and future directions, aiming to provide a reference for clinical practice and related research.

关键词

人工智能 / 诊断 / 治疗 / 预测模型 / 机器学习 / 早产儿

Key words

Artificial intelligence / Diagnosis / Treatment / Predictive model / Machine learning / Preterm infant

引用本文

引用格式 ▾
袁英,唐翎瀚,管利荣. 人工智能在早产儿疾病诊疗中的研究进展[J]. 中国当代儿科杂志, 2026, 28(05): 629-635 DOI:10.7499/j.issn.1008-8830.2507030

登录浏览全文

4963

注册一个新账户 忘记密码

早产是新生儿死亡的重要危险因素,并会带来长期的身体、神经发育及社会经济等方面的负面影响1。早产儿因其器官发育不成熟,易并发多种疾病,包括呼吸窘迫综合征(respiratory distress syndrome, RDS)、支气管肺发育不良(bronchopulmonary dysplasia, BPD)、早产儿视网膜病变(retinopathy of prematurity, ROP)、坏死性小肠结肠炎(necrotizing enterocolitis, NEC)及脑损伤等,给临床诊疗带来巨大挑战2。这些疾病的早期预测和精准诊断对改善早产儿预后至关重要,但传统方法往往依赖临床经验和主观判断,难以实现早期预警和个体化干预。
人工智能(artificial intelligence, AI)在医学影像分析、生理信号监测和临床决策支持等方面的应用取得了显著进展,展现出巨大的潜力3-5。目前,临床上常用的AI技术包括机器学习、深度学习和大模型等(表1)。AI技术的快速发展为早产儿疾病的诊疗提供了新的解决方案6。本文聚焦于AI在RDS、BPD、ROP、NEC等早产儿常见疾病中的创新应用,以期为相关临床诊疗提供参考。

1 AI在早产儿疾病预测中的应用

1.1 AI在早产预测中的应用

AI在早产风险预测领域已成为研究热点。Chakoory等7研究分析来自561名孕妇的1 290个阴道样本的宏基因组数据,构建并训练深度神经网络(deep neural network, DNN),用于预测足月和早产分娩,模型预测准确率为84.10%,受试者操作特征曲线下面积(area under the receiver operating characteristic curve, AUROC)为0.875±0.110。Zhang等8采用机器学习(machine learning, ML)算法构建早产预测模型,发现基于电子健康记录(electronic health record, EHR)的AdaBoost模型对“非早产”的预测准确率达100%,对“早产”的预测准确率为72.73%。AI技术在早产预测中表现出良好的应用价值,可为临床提供有效的早产风险评估工具,对优化临床决策、改善预后具有重要意义。

1.2 早产儿并发症的早期预警系统

韩国一项多中心研究纳入13 087例极低出生体重早产儿,基于婴儿基本信息、孕母病史、复苏过程及出生后检测数据,比较7种ML模型对RDS的预测效能,并构建5层DNN模型以提升预测效果,该模型灵敏度为83.03%,特异度为87.50%,准确率为84.07%,平衡准确率为85.26%,AUROC达0.91879。该AI模型可辅助临床做好新生儿复苏准备,减少肺表面活性物质的不合理使用,降低无RDS早产儿发生医源性损伤的风险。

Moreira等10分析97例极低出生体重儿全血微阵列数据,采用含5种基因组合的ML模型预测BPD,曲线下面积(area under the curve, AUC)范围为85.8%~96.1%,其中极端梯度提升树模型AUC达96.1%(95%CI:89.7%~100%),重要基因依次为PNPOMSANTD2CD4SNX1及P2RX7。Chen等11利用ML方法构建早产儿ROP风险预测模型,将出生体重、胎龄、性别、是否多胎分娩及分娩方式等风险因素输入逻辑回归、决策树及多层感知器等算法,实现对ROP及需治疗ROP的风险预测。Sylvester等12构建多变量NEC进展预测模型,采用27项临床参数联合尿生物标志物(纤维蛋白原肽FGA1826、FGA1883及FGA2659)的集成模型,可正确预测所有测试病例的NEC结局。

AI在早产儿并发症早期预警系统中发挥关键作用,可通过各类预测模型快速识别异常,预测RDS、BPD、ROP、NEC等高风险疾病,其高精度分析能力可降低误诊率,为临床决策提供辅助。

1.3 多模态数据融合在预测中的应用

Routier等13研究探讨多模态模型,通过整合影像学、代谢组学及临床数据预测极早产儿死亡或神经发育障碍风险,结果显示该多模态模型AUC达91.7%(95%CI:86.4%~97.0%)显著高于单模态模型(P<0.003)。此外,多模态数据融合也应用于NEC风险预测,其AUROC为0.8,具有较高的预测准确性14。未来,随着物联网技术的普及,整合实时多模态监测数据与AI分析技术,可进一步提升对早产儿各类并发症的预测精准度。

2 AI在早产儿疾病诊断中的进展

2.1 医学影像分析

AI在早产儿医学影像分析领域展现出显著优势,尤其在颅脑超声、胸部X线等影像的自动化解读方面。以颅脑超声为例,研究显示,深度学习模型可有效区分正常与异常颅脑超声图像并快速解读扫描结果,其AUROC为0.86(95%CI:0.82~0.90),精度-召回AUC为0.87(95%CI:0.84~0.90)15。Zivojinovic等16采用卷积神经网络分析并分类早产儿脑超声图像的密度差异,实现对新生儿缺氧缺血性脑病的快速、准确识别,这对及时采取适宜治疗措施、改善患儿长期预后至关重要。

2.2 生理信号监测

在生理信号监测领域,AI可通过解析早产儿心电、血压等时序数据实现疾病早期预警,显著提升医生对生理信号异常模式的识别效率。Park等17基于409例早产儿的EHR,通过AI分析心率、呼吸、血压等动态参数可提前3.3 d预警动脉导管未闭,准确率为84%。Leon等18利用心率变异性及其新特征——可见性图索引,通过ML算法预测早产儿晚发性败血症,为临床提供了一种潜在的非侵入性、实时监测手段。

2.3 生物标志物分析与诊断模型

AI在生物标志物多维度分析中表现出强大的模式识别能力。Wang等19采用高通量液相色谱法对62例脑性瘫痪儿童和60例健康儿童对照的血浆样本进行氨基酸代谢组学分析,结果显示代谢谱呈特征性改变,尤其在早产儿脑性瘫痪中,β⁃氨基异丁酸、色氨酸及牛磺酸水平降低,三者联合可作为潜在的生物标志物组合用于辅助诊断与风险预测,其中早产儿脑性瘫痪的判别模型AUC达0.874(95%CI:0.732~1.000)。Lin等20通过对早产儿粪便细菌DNA测序获取肠道菌群数据,构建新型可解释的多次实例学习系统,可有效预测NEC的发生风险。

3 AI在早产儿治疗中的创新应用

3.1 个性化治疗方案推荐

AI可整合多模态临床数据建立预测模型,为不同风险分层的早产儿提供个性化干预策略。基于深度学习衍生的血管严重程度评分应用于ROP临床管理,可提高视网膜病变诊断的一致性,通过AI算法评估提供针对性治疗建议,有效优化医疗资源配置21-22。AI还可结合连续血糖监测数据,通过CLAIR风险指数预测脑室内出血风险,其灵敏度达100%,为早期个性化治疗提供依据23。Xu等24构建的XGBoost模型可预测早产儿喂养不耐受的风险因素,准确率高达87.62%,可用于临床指导个性化喂养方案制定。

3.2 呼吸机参数优化

传统呼吸机参数调整依赖临床经验,而AI系统可实时分析血氧饱和度、呼吸波形、血气分析等动态监测数据,优化呼吸机参数2527。此外,AI算法可实时评估患者血流动力学状态,协助临床提高警惕性,把握早期治疗时机,进而缩短机械通气时间2829

3.3 药物剂量计算与不良反应预测

早产儿药代动力学特性复杂,传统给药方案易导致药物蓄积或疗效不足30。van den Anker等31研究表明,通过整合群体药代动力学模型与个体化生理参数,可优化个体化给药方案,实现安全有效用药。在新生儿疼痛管理方面,AI系统可综合面部表情识别、哭声频率分析及生命体征变化,预测术后疼痛发作的时间窗,为超前镇痛提供决策支持32。一项前瞻性队列研究显示,利用ML技术对新生儿啼哭进行声学分析,可助力改善新生儿戒断综合征的评估、诊断和管理,同时推动此类患儿治疗方案的标准化33

4 AI在早产儿预后评估中的作用

4.1 神经发育结局预测

AI在早产儿神经发育结局预测方面具有潜在应用价值。Bowe等34通过构建逻辑回归模型,纳入社会人口统计学、临床信息等90个变量用于预测极早产儿认知延迟情况,该模型AUC为0.77。此外,另有研究采用图卷积网络模型预测早产儿认知、语言、运动技能等方面的神经发育缺陷,AUC达0.72~0.7535。研究表明,ML在早产儿神经发育结局预测中已取得初步成功,但仍面临诸多挑战,包括针对神经发育结局的最具预示力的临床和大脑特征尚未达成共识,以及ML技术仍需进一步探索完善36

4.2 长期健康风险评估与随访管理

De Francesco等37利用EHR开发纵向风险评估模型,通过综合关联分析发现多种已知的母亲及新生儿特征与特定新生儿结局的关联,为新生儿结局的探索和预测提供了重要资源。ML可对足月儿及晚期早产儿的发育风险进行分级,该研究采用分类和回归树算法,可针对最可能获益的目标人群扩大高危婴儿随访登记范围,有利于临床随访管理与进一步干预38

5 AI技术的挑战与局限性

5.1 AI辅助诊断早产儿疾病的局限性

尽管AI在早产儿诊疗中的应用已取得一定进展,但在预测准确性等方面仍存在不足。例如,在预测早产儿拔管成功方面,目前AI模型的效能尚未超越临床预测指标39。Demirci等40研究显示,虽可通过ML模型结合围产期变量预测早产儿25月龄时的发育迟缓,但模型特异度较低,存在过度预测风险。

5.2 数据质量与标准化问题

AI在医疗领域的应用高度依赖高质量、标准化的数据,但当前医疗数据普遍存在碎片化、异构性及标注不一致等问题41。例如,在新生儿血压监测研究中,非侵入式传感器采集的数据易受运动伪影和环境干扰影响,导致模型训练效果受限42。此外,资源匮乏地区因缺乏代表性数据,导致AI算法存在地域性偏差43。数据标准化需建立跨机构协作框架、统一数据采集协议、质量控制流程及伦理审查机制,这对实现AI技术的临床转化至关重要44

5.3 模型可解释性与临床接受度

当前深度学习模型的“黑箱”特性严重制约了其在临床的落地应用45。现有研究中,多数未开展独立的模型测试或缺乏有效的对照组,导致模型的真实效能和临床价值难以评估46。AI模型的可解释性技术旨在揭示复杂模型内部的决策过程,使用户能够理解和信任模型的预测结果,主要分为两大类:一类是模型内在解释性设计,如决策树和线性模型等简单的模型;另一类是黑盒模型的事后解释技术,如SHAP、LIME、Grad⁃CAM等。在早产儿脑损伤风险预测模型中,通过SHAP值分析,可识别机械通气、体重、贫血等关键风险因素47。未来,需开发基于注意力机制、特征重要性排序的可解释性AI技术,同时通过多中心临床试验验证模型决策与临床结局的因果关系48

5.4 伦理与隐私问题

医疗AI的隐私保护与数据安全面临重大挑战。当前多数AI技术由私人实体掌控,这可能导致对患者健康信息的获取、使用和保护超出常规范畴,易引发隐私问题49。例如,ChatGPT、Google Bard等AI聊天机器人依托AI和自然语言处理技术生成回复,其模型基于数十亿个数据点训练而成,这意味着此类模型可能在未经授权的情况下访问大量用户数据,其中或包括患者敏感信息,从而加剧数据安全风险50。数据泄露不仅会侵犯患者隐私权,还可能引发严重的法律和伦理后果。此外,AI医疗事故的责任划分复杂,目前多数国家和地区的法律体系尚未针对此类事故形成统一的责任归属标准。

6 小结与展望

AI在早产儿疾病诊疗中的应用体现了医学技术的现代化与智能化发展趋势,为新生儿重症监护领域带来了革命性变革。现有研究表明,AI在提高诊断准确性、优化个体化治疗方案及预测疾病预后等方面优势显著922。此外,AI驱动的决策支持系统可提供患者特定信息和循证建议,帮助临床医师提高诊断准确性、优化治疗方案,同时减少医疗差错51

尽管AI技术在早产儿疾病诊疗领域前景广阔,但实际应用中仍面临诸多挑战。数据质量是影响AI模型性能的关键因素,早产儿临床数据常存在样本量不足、标注不一致及偏倚等问题,这可能导致模型泛化能力受限。AI模型可解释性不足,在高风险诊疗决策中易因“黑箱”问题引发信任危机,从而影响临床接受度。此外,数据隐私、算法公平性及辅助决策责任归属等伦理问题也需重点关注。

针对上述问题可开展针对性研究。一方面,需从多维度提升AI模型性能,包括提高数据质量、规范数据处理流程、制定行业标准及优化算法设计。具体可通过推动多中心协作和数据共享扩大样本量,构建更具代表性和多样性的数据集;制定统一的数据采集与处理规范;改进模型设计,采用混合架构,引入反向注意力机制,并通过优化超参数及数据预处理与增强策略提升模型性能,从而增加临床应用的可靠性。另一方面,严格的临床验证必不可少,通过前瞻性研究评估AI工具在真实医疗场景中的有效性和安全性。同时,加强数据隐私保护,制定相关伦理准则与规范,建立严格的数据安全存储和监管机制并定期核查;建立AI医疗事故责任归属标准,构建多方共担、责权清晰的责任体系。此外,AI技术需与现有医疗流程深度整合,避免加重医护人员工作负担。AI在早产儿诊疗中的应用不仅需要技术创新,还需要政策支持、伦理体系完善及医患双方共同参与。

综上所述,AI在早产儿疾病诊疗中具有巨大潜力,但其深度临床应用需建立在科学、伦理与临床实践的多维平衡之上。通过跨学科协作和持续优化,AI有望成为改善早产儿健康结局的重要工具,推动新生儿医学迈向智能化、精准化的新时代,最终实现AI在早产儿疾病诊疗领域的广泛应用,为早产儿提供高质量的医疗服务。

参考文献

[1]

Ohuma EO, Moller AB, Bradley E, et al. National, regional, and global estimates of preterm birth in 2020, with trends from 2010: a systematic analysis[J]. Lancet, 2023, 402(10409): 1261-1271. DOI: 10.1016/S0140-6736(23)00878-4 .

[2]

Humberg A, Fortmann I, Siller B, et al. Preterm birth and sustained inflammation: consequences for the neonate[J]. Semin Immunopathol, 2020, 42(4): 451-468. PMCID: PMC7508934. DOI: 10.1007/s00281-020-00803-2 .

[3]

Li X, Zhao L, Zhang L, et al. Artificial general intelligence for medical imaging analysis[J]. IEEE Rev Biomed Eng, 2025, 18: 113-129. DOI: 10.1109/RBME.2024.3493775 .

[4]

Sundrani S, Chen J, Jin BT, et al. Predicting patient decompensation from continuous physiologic monitoring in the emergency department[J]. NPJ Digit Med, 2023, 6(1): 60. PMCID: PMC10073111. DOI: 10.1038/s41746-023-00803-0 .

[5]

Verlingue L, Boyer C, Olgiati L, et al. Artificial intelligence in oncology: ensuring safe and effective integration of language models in clinical practice[J]. Lancet Reg Health Eur, 2024, 46: 101064. PMCID: PMC11406067. DOI: 10.1016/j.lanepe.2024.101064 .

[6]

Shu CH, Zebda R, Espinosa C, et al. Early prediction of mortality and morbidities in VLBW preterm neonates using machine learning[J]. Pediatr Res, 2025, 97(6): 2056-2064. PMCID: PMC12249422. DOI: 10.1038/s41390-024-03604-7 .

[7]

Chakoory O, Barra V, Rochette E, et al. DeepMPTB: a vaginal microbiome-based deep neural network as artificial intelligence strategy for efficient preterm birth prediction[J]. Biomark Res, 2024, 12(1): 25. PMCID: PMC10865581. DOI: 10.1186/s40364-024-00557-1 .

[8]

Zhang Y, Du S, Hu T, et al. Establishment of a model for predicting preterm birth based on the machine learning algorithm[J]. BMC Pregnancy Childbirth, 2023, 23(1): 779. PMCID: PMC10636958. DOI: 10.1186/s12884-023-06058-7 .

[9]

Jang W, Choi YS, Kim JY, et al. Artificial intelligence: driven respiratory distress syndrome prediction for very low birth weight infants: Korean multicenter prospective cohort study[J]. J Med Internet Res, 2023, 25: e47612. PMCID: PMC10366668. DOI: 10.2196/47612 .

[10]

Moreira A, Tovar M, Smith AM, et al. Development of a peripheral blood transcriptomic gene signature to predict bronchopulmonary dysplasia[J]. Am J Physiol Lung Cell Mol Physiol, 2023, 324(1): L76-L87. PMCID: PMC9829478. DOI: 10.1152/ajplung.00250.2022 .

[11]

Chen S, Zhao X, Wu Z, et al. Multi-risk factors joint prediction model for risk prediction of retinopathy of prematurity[J]. EPMA J, 2024, 15(2): 261-274. PMCID: PMC11147992. DOI: 10.1007/s13167-024-00363-7 .

[12]

Sylvester KG, Ling XB, Liu GY, et al. A novel urine peptide biomarker-based algorithm for the prognosis of necrotising enterocolitis in human infants[J]. Gut, 2014, 63(8): 1284-1292. PMCID: PMC4161026. DOI: 10.1136/gutjnl-2013-305130 .

[13]

Routier L, Querne L, Ghostine-Ramadan G, et al. Predicting the neurodevelopmental outcome in extremely preterm newborns using a multimodal prognostic model including brain function information[J]. JAMA Netw Open, 2023, 6(3): e231590. PMCID: PMC9996404. DOI: 10.1001/jamanetworkopen.2023.1590 .

[14]

Qi G, Huang S, Lai D, et al. An improved joint non-negative matrix factorization for identifying surgical treatment timing of neonatal necrotizing enterocolitis[J]. Bosn J Basic Med Sci, 2022, 22(6): 972-981. PMCID: PMC9589314. DOI: 10.17305/bjbms.2022.7046 .

[15]

Ahmad T, Guida A, Stewart S, et al. Can deep learning classify cerebral ultrasound images for the detection of brain injury in very preterm infants?[J]. Eur Radiol, 2025, 35(4): 1948-1958. DOI: 10.1007/s00330-024-11028-4 .

[16]

Zivojinovic S, Petrovic Savic S, Prodanovic T, et al. Neurosonographic classification in premature infants receiving omega-3 supplementation using convolutional neural networks[J]. Diagnostics (Basel), 2024, 14(13): 1342. PMCID: PMC11241385. DOI: 10.3390/diagnostics14131342 .

[17]

Park S, Moon J, Eun H, et al. Artificial intelligence-based diagnostic support system for patent ductus arteriosus in premature infants[J]. J Clin Med, 2024, 13(7): 2089. PMCID: PMC11012712. DOI: 10.3390/jcm13072089 .

[18]

Leon C, Carrault G, Pladys P, et al. Early detection of late onset sepsis in premature infants using visibility graph analysis of heart rate variability[J]. IEEE J Biomed Health Inform, 2021, 25(4): 1006-1017. DOI: 10.1109/JBHI.2020.3021662 .

[19]

Wang D, Song J, Cheng Y, et al. Targeting the metabolic profile of amino acids to identify the key metabolic characteristics in cerebral palsy[J]. Front Mol Neurosci, 2023, 16: 1237745. PMCID: PMC10470834. DOI: 10.3389/fnmol.2023.1237745 .

[20]

Lin YC, Salleb-Aouissi A, Hooven TA. Interpretable prediction of necrotizing enterocolitis from machine learning analysis of premature infant stool microbiota[J]. BMC Bioinformatics, 2022, 23(1): 104. PMCID: PMC8953333. DOI: 10.1186/s12859-022-04618-w .

[21]

Campbell JP, Kim SJ, Brown JM, et al. Evaluation of a deep learning-derived quantitative retinopathy of prematurity severity scale[J]. Ophthalmology, 2021, 128(7): 1070-1076. PMCID: PMC8076329. DOI: 10.1016/j.ophtha.2020.10.025 .

[22]

deCampos-Stairiker MA, Coyner AS, Gupta A, et al. Epidemiologic evaluation of retinopathy of prematurity severity in a large telemedicine program in India using artificial intelligence[J]. Ophthalmology, 2023, 130(8): 837-843. PMCID: PMC10524227. DOI: 10.1016/j.ophtha.2023.03.026 .

[23]

Galderisi A, Zammataro L, Losiouk E, et al. Continuous glucose monitoring linked to an artificial intelligence risk index: early footprints of intraventricular hemorrhage in preterm neonates[J]. Diabetes Technol Ther, 2019, 21(3): 146-153. DOI: 10.1089/dia.2018.0383 .

[24]

Xu H, Peng X, Peng Z, et al. Construction and SHAP interpretability analysis of a risk prediction model for feeding intolerance in preterm newborns based on machine learning[J]. BMC Med Inform Decis Mak, 2024, 24(1): 342. PMCID: PMC11572196. DOI: 10.1186/s12911-024-02751-5 .

[25]

Zhu S, Liu S, Jing X, et al. Innovative approaches in imaging photoplethysmography for remote blood oxygen monitoring[J]. Sci Rep, 2024, 14(1): 19144. PMCID: PMC11333616. DOI: 10.1038/s41598-024-70192-1 .

[26]

Bondala VR, Komalla AR. An efficient model for extracting respiratory and blood oxygen saturation data from photoplethysmogram signals by removing motion artifacts using heuristic-aided ensemble learning model[J]. Comput Biol Med, 2024, 180: 108911. DOI: 10.1016/j.compbiomed.2024.108911 .

[27]

Hu J, Han Z, Heidari AA, et al. Detection of COVID-19 severity using blood gas analysis parameters and Harris hawks optimized extreme learning machine[J]. Comput Biol Med, 2022, 142: 105166. PMCID: PMC8701842. DOI: 10.1016/j.compbiomed.2021.105166 .

[28]

Rahman A, Chang Y, Dong J, et al. Early prediction of hemodynamic interventions in the intensive care unit using machine learning[J]. Crit Care, 2021, 25(1): 388. PMCID: PMC8590869. DOI: 10.1186/s13054-021-03808-x .

[29]

Liao KM, Ko SC, Liu CF, et al. Development of an interactive AI system for the optimal timing prediction of successful weaning from mechanical ventilation for patients in respiratory care centers[J]. Diagnostics (Basel), 2022, 12(4): 975. PMCID: PMC9030191. DOI: 10.3390/diagnostics12040975 .

[30]

Zhang W, Zhang Q, Cao Z, et al. Physiologically based pharmacokinetic modeling in neonates: current status and future perspectives[J]. Pharmaceutics, 2023, 15(12): 2765. PMCID: PMC10747965. DOI: 10.3390/pharmaceutics15122765 .

[31]

van den Anker J, Allegaert K. Considerations for drug dosing in premature infants[J]. J Clin Pharmacol, 2021, 61 (): S141-S151. DOI: 10.1002/jcph.1884 .

[32]

Salekin MS, Mouton PR, Zamzmi G, et al. Future roles of artificial intelligence in early pain management of newborns[J]. Paediatr Neonatal Pain, 2021, 3(3): 134-145. PMCID: PMC8975206. DOI: 10.1002/pne2.12060 .

[33]

Manigault AW, Sheinkopf SJ, Silverman HF, et al. Newborn cry acoustics in the assessment of neonatal opioid withdrawal syndrome using machine learning[J]. JAMA Netw Open, 2022, 5(10): e2238783. PMCID: PMC9614579. DOI: 10.1001/jamanetworkopen.2022.38783 .

[34]

Bowe AK, Lightbody G, Staines A, et al. Prediction of 2-year cognitive outcomes in very preterm infants using machine learning methods[J]. JAMA Netw Open, 2023, 6(12): e2349111. PMCID: PMC10751596. DOI: 10.1001/jamanetworkopen.2023.49111 .

[35]

Li H, Wang J, Li Z, et al. Supervised contrastive learning enhances graph convolutional networks for predicting neurodevelopmental deficits in very preterm infants using brain structural connectome[J]. Neuroimage, 2024, 291: 120579. PMCID: PMC11059107. DOI: 10.1016/j.neuroimage.2024.120579 .

[36]

Baker S, Kandasamy Y. Machine learning for understanding and predicting neurodevelopmental outcomes in premature infants: a systematic review[J]. Pediatr Res, 2023, 93(2): 293-299. PMCID: PMC9153218. DOI: 10.1038/s41390-022-02120-w .

[37]

De Francesco D, Reiss JD, Roger J, et al. Data-driven longitudinal characterization of neonatal health and morbidity[J]. Sci Transl Med, 2023, 15(683): eadc9854. PMCID: PMC10197092. DOI: 10.1126/scitranslmed.adc9854 .

[38]

Carlton K, Zhang J, Cabacungan E, et al. Machine learning risk stratification for high-risk infant follow-up of term and late preterm infants[J]. Pediatr Res. PMID: PMCID: PMC11669732. DOI: 10.1038/s41390-024-03338-6 . Epub ahead of print.

[39]

Jenkinson AC, Dassios T, Greenough A. Artificial intelligence in the NICU to predict extubation success in prematurely born infants[J]. J Perinat Med, 2024, 52(2): 119-125. DOI: 10.1515/jpm-2023-0454 .

[40]

Demirci GM, Kittler PM, Phan HTT, et al. Predicting mental and psychomotor delay in very pre-term infants using machine learning[J]. Pediatr Res, 2024, 95(3): 668-678. PMCID: PMC10899098. DOI: 10.1038/s41390-023-02713-z .

[41]

Madni HA, Umer RM, Foresti GL. Robust federated learning for heterogeneous model and data[J]. Int J Neural Syst, 2024, 34(4): 2450019. DOI: 10.1142/S0129065724500199 .

[42]

Baker S, Yogavijayan T, Kandasamy Y. Towards non-invasive and continuous blood pressure monitoring in neonatal intensive care using artificial intelligence: a narrative review[J]. Healthcare (Basel), 2023, 11(24): 3107. PMCID: PMC10743031. DOI: 10.3390/healthcare11243107 .

[43]

Kim S, Fischetti C, Guy M, et al. Artificial intelligence (AI) applications for point of care ultrasound (POCUS) in low-resource settings: a scoping review[J]. Diagnostics (Basel), 2024, 14(15): 1669. PMCID: PMC11312308. DOI: 10.3390/diagnostics14151669 .

[44]

Alberto IRI, Alberto NRI, Ghosh AK, et al. The impact of commercial health datasets on medical research and health-care algorithms[J]. Lancet Digit Health, 2023, 5(5): e288-e294. PMCID: PMC10155113. DOI: 10.1016/S2589-7500(23)00025-0 .

[45]

Karako K, Tang W. Applications of and issues with machine learning in medicine: bridging the gap with explainable AI[J]. Biosci Trends, 2025, 18(6): 497-504. DOI: 10.5582/bst.2024.01342 .

[46]

Krarup MMK, Krokos G, Subesinghe M, et al. Artificial intelligence for the characterization of pulmonary nodules, lung tumors and mediastinal nodes on PET/CT[J]. Semin Nucl Med, 2021, 51(2): 143-156. DOI: 10.1053/j.semnuclmed.2020.09.001 .

[47]

He Z, Zhang R, Qu P, et al. Development and validation of an explainable model of brain injury in premature infants: a prospective cohort study[J]. Comput Methods Programs Biomed, 2025, 260: 108559. DOI: 10.1016/j.cmpb.2024.108559 .

[48]

Di Martino F, Delmastro F. Explainable AI for clinical and remote health applications: a survey on tabular and time series data[J]. Artif Intell Rev, 2023, 56(6): 5261-5315. PMCID: PMC9607788. DOI: 10.1007/s10462-022-10304-3 .

[49]

Murdoch B. Privacy and artificial intelligence: challenges for protecting health information in a new era[J]. BMC Med Ethics, 2021, 22(1): 122. PMCID: PMC8442400. DOI: 10.1186/s12910-021-00687-3 .

[50]

Li J. Security implications of AI chatbots in health care[J]. J Med Internet Res, 2023, 25: e47551. PMCID: PMC10716748. DOI: 10.2196/47551 .

[51]

Ouanes K, Farhah N. Effectiveness of artificial intelligence (AI) in clinical decision support systems and care delivery[J]. J Med Syst, 2024, 48(1): 74. DOI: 10.1007/s10916-024-02098-4 .

RIGHTS & PERMISSIONS

版权所有 © 2023中国当代儿科杂志

AI Summary AI Mindmap
PDF (550KB)

122

访问

0

被引

详细

导航
相关文章

AI思维导图

/