人工智能在注意缺陷多动障碍儿童医学应用中的研究进展

doi:10.12114/j.issn.1007-9572.2025.0229

中国全科医学 ›› 2026, Vol. 29 ›› Issue (36): 5466-5472.DOI: 10.12114/j.issn.1007-9572.2025.0229

• 综述与专论 • 上一篇

人工智能在注意缺陷多动障碍儿童医学应用中的研究进展

吴莉¹, 王浩懿², 邱向阳³, 袁静⁴, 陈如男¹^,^*()

1．200433　上海市，海军军医大学护理系
2．200433　上海市，海军军医大学基础医学院
3．200032　上海市，复旦大学营养研究院
4．200433　上海市，海军军医大学第一附属医院儿科

收稿日期:2025-07-25 修回日期:2025-09-01 出版日期:2026-12-20 发布日期:2026-10-30
通讯作者: 陈如男
作者贡献：
吴莉提出研究思路，设计研究方案；负责论文起草和撰写；王浩懿负责文献检索、收集，设计研究方案，负责论文修改；邱向阳提出研究思路，负责文献检索、采集，负责图表制作；袁静负责提供论文临床上的指导建议、论文修订意见；陈如男负责论文全程修改意见和最终版本修订，对论文负责。
基金资助:
上海市卫生健康委员会科研项目(202140010); 海军军医大学社科培育基金(2022SK023)

Progress of Artificial Intelligence in Children with Attention Deficit Hyperactivity Disorder

WU Li¹, WANG Haoyi², QIU Xiangyang³, YUAN Jing⁴, CHEN Runan¹^,^*()

1. Department of Nursing, Naval Medical University, Shanghai 200433, China
2. School of Basic Medical Sciences, Naval Medical University, Shanghai 200433, China
3. Institute of Nutrition, Fudan University, Shanghai 200032, China
4. Department of Pediatrics, First Affiliated Hospital of Navy Medical University, Shanghai 200433, China

Received:2025-07-25 Revised:2025-09-01 Published:2026-12-20 Online:2026-10-30
Contact: CHEN Runan

摘要/Abstract

摘要： 注意缺陷多动障碍（ADHD）是一种严重影响儿童生长发育的神经障碍性疾病，准确高效的诊断和干预策略是促进患儿身心健康和提高生活质量的关键。当前，临床ADHD诊疗面临诊断主观性强、疗效个体差异大、共病复杂等诸多挑战。人工智能（AI）凭借其强大的算法分析和精准的特征提取能力在ADHD儿童的早期诊断和疗效评估等方面展现出巨大潜力，AI技术能够提供多维度临床决策支持，评估其预后疗效。本文就近年来AI在ADHD疾病诊断、症状和药物疗效预测以及跨诊断方面的研究进展进行总结，同时提炼核心共性挑战，提出针对性建议，为促进AI技术在ADHD儿童医学中的发展应用提供参考。

关键词: 人工智能, 注意缺陷多动障碍, 辅助诊断, 疾病特征预测, 药物疗效预测

Abstract:

Attention deficit hyperactivity disorder (ADHD) is a neurological disorder that significantly affects children's growth and development. Accurate and efficient diagnosis and intervention strategies are crucial for promoting the physical and mental health of affected children and improving their quality of life. Currently, clinical diagnosis and treatment of ADHD face numerous challenges, including strong subjectivity in diagnosis, significant individual differences in treatment efficacy, and complex comorbidities. Artificial intelligence (AI), with its powerful algorithmic analysis and precise feature extraction capabilities, has shown great potential in the early diagnosis and efficacy evaluation of ADHD in children. It is expected to provide multi-dimensional clinical decision support and assess prognosis through AI technology. This article summarizes the research progress of AI in ADHD diagnosis, symptom and drug efficacy prediction, and cross-diagnosis in recent years. At the same time, it distills core common challenges and puts forward targeted suggestions to provide references for promoting the development and application of AI technology in ADHD.

Key words: Artificial intelligence, Attention deficit hyperactivity disorder, Auxiliary diagnosis, Prediction of disease characteristics, Drug efficacy prediction

中图分类号:

R 542.22　R 364.5

吴莉,王浩懿,邱向阳,等. 人工智能在注意缺陷多动障碍儿童医学应用中的研究进展[J]. 中国全科医学, 2026, 29(36): 5466-5472. DOI: 10.12114/j.issn.1007-9572.2025.0229.
WU Li,WANG Haoyi,QIU Xiangyang, et al. Progress of Artificial Intelligence in Children with Attention Deficit Hyperactivity Disorder[J]. Chinese General Practice, 2026, 29(36): 5466-5472. DOI: 10.12114/j.issn.1007-9572.2025.0229.

图/表 1

参考文献 42

[1]	Ayano G, Demelash S, Gizachew Y, et al. The global prevalence of attention deficit hyperactivity disorder in children and adolescents: an umbrella review of meta-analyses[J]. J Affect Disord, 2023, 339: 860-866. DOI: 10.1016/j.jad.2023.07.071.
[2]	Wang T T, Liu K H, Li Z Z, et al. Prevalence of attention deficit/hyperactivity disorder among children and adolescents in China: a systematic review and meta-analysis[J]. BMC Psychiatry, 2017, 17(1): 32. DOI: 10.1186/s12888-016-1187-9.
[3]	Sibley M H, Swanson J M, Arnold L E, et al. Defining ADHD symptom persistence in adulthood: optimizing sensitivity and specificity[J]. J Child Psychol Psychiatry, 2017, 58(6): 655-662. DOI: 10.1111/jcpp.12620.
[4]	陈亚敏, 程露颖, 郑卿勇, 等. 人工智能与神经科学研究主题分析[J]. 兰州大学学报(医学版), 2022, 48(3): 59-65. DOI: 10.13885/j.issn.1000-2812.2022.03.013.
[5]	周小芹, 刘慧珍, 王婷, 等. 人工智能赋能医学领域的挑战与发展方向[J]. 中国胸心血管外科临床杂志, 2025, 32(2): 244-251.
[6]	Chen I C, Chang C L, Chang M H, et al. The utility of wearable electroencephalography combined with behavioral measures to establish a practical multi-domain model for facilitating the diagnosis of young children with attention-deficit/hyperactivity disorder[J]. J Neurodev Disord, 2024, 16(1): 62. DOI: 10.1186/s11689-024-09578-1.
[7]	Fang H, Guo J J, Wu H. Wearable triboelectric devices for haptic perception and VR/AR applications[J]. Nano Energy, 2022, 96: 107112. DOI: 10.1016/j.nanoen.2022.107112.
[8]	Kim W P, Kim H J, Pack S P, et al. Machine learning-based prediction of attention-deficit/hyperactivity disorder and sleep problems with wearable data in children[J]. JAMA Netw Open, 2023, 6(3): e233502. DOI: 10.1001/jamanetworkopen.2023.3502.
[9]	Muñoz-Organero M, Powell L, Heller B, et al. Automatic extraction and detection of characteristic movement patterns in children with ADHD based on a convolutional neural network(CNN)and acceleration images[J]. Sensors, 2018, 18(11): 3924. DOI: 10.3390/s18113924.
[10]	De Hond A A H, Leeuwenberg A M, Hooft L, et al. Guidelines and quality criteria for artificial intelligence-based prediction models in healthcare: a scoping review[J]. NPJ Digit Med, 2022, 5(1): 2. DOI: 10.1038/s41746-021-00549-7.
[11]	Collins G S, Moons K G M. Reporting of artificial intelligence prediction models[J]. Lancet, 2019, 393(10181): 1577-1579. DOI: 10.1016/S0140-6736(19)30037-6.
[12]	Riley R D, Ensor J, Snell K I E, et al. Calculating the sample size required for developing a clinical prediction model[J]. BMJ, 2020, 368: m441. DOI: 10.1136/bmj.m441.
[13]	张康福茜, 曹庆久. 儿童注意缺陷多动障碍脑磁共振研究进展[J]. 中国实用儿科杂志, 2023, 38(8): 575-579. DOI: 10.19538/j.ek2023080604.
[14]	Alnazer I, Bourdon P, Urruty T, et al. Recent advances in medical image processing for the evaluation of chronic kidney disease[J]. Med Image Anal, 2021, 69: 101960. DOI: 10.1016/j.media.2021.101960.
[15]	朱莉, 张丽英, 韩云涛, 等. 基于卷积神经网络的注意缺陷多动障碍分类研究[J]. 生物医学工程学杂志, 2017, 34(1): 99-105.
[16]	Ke H J, Wang F Q, Ma H Y, et al. ADHD identification and its interpretation of functional connectivity using deep self-attention factorization[J]. Knowl Based Syst, 2022, 250: 109082. DOI: 10.1016/j.knosys.2022.109082.
[17]	陈丽舟, 李倩, 罗乐凯, 等. 不同亚型注意缺陷多动障碍患者的脑白质微观结构差异及分类研究[J]. 华西医学, 2023, 38(3): 408-415.
[18]	Ahire N, Awale R N, Wagh A. Electroencephalogram(EEG)based prediction of attention deficit hyperactivity disorder(ADHD)using machine learning[J]. Appl Neuropsychol Adult, 2025, 32(4): 966-977. DOI: 10.1080/23279095.2023.2247702.
[19]	徐琴芳, 王延培. 脑电测量在儿童发展性障碍诊断中的应用[J]. 心理科学进展, 2017, 25(12): 2136-2144.
[20]	García-Ponsoda S, Maté A, Trujillo J. Refining ADHD diagnosis with EEG: the impact of preprocessing and temporal segmentation on classification accuracy[J]. Comput Biol Med, 2024, 183: 109305. DOI: 10.1016/j.compbiomed.2024.109305.
[21]	Vahid A, Bluschke A, Roessner V, et al. Deep learning based on event-related EEG differentiates children with ADHD from healthy controls[J]. J Clin Med, 2019, 8(7): 1055. DOI: 10.3390/jcm8071055.
[22]	Sanchis J, García-Ponsoda S, Teruel M A, et al. A novel approach to identify the brain regions that best classify ADHD by means of EEG and deep learning[J]. Heliyon, 2024, 10(4): e26028. DOI: 10.1016/j.heliyon.2024.e26028.
[23]	Khare S K, Acharya U R. An explainable and interpretable model for attention deficit hyperactivity disorder in children using EEG signals[J]. Comput Biol Med, 2023, 155: 106676. DOI: 10.1016/j.compbiomed.2023.106676.
[24]	Collins G S, Moons K G M, Dhiman P, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods[J]. BMJ, 2024, 385: q902. DOI: 10.1136/bmj.q902.
[25]	Christodoulou E, Ma J, Collins G S, et al. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models[J]. J Clin Epidemiol, 2019, 110: 12-22. DOI: 10.1016/j.jclinepi.2019.02.004.
[26]	Slobodin O, Yahav I, Berger I. A machine-based prediction model of ADHD using CPT data[J]. Front Hum Neurosci, 2020, 14: 560021. DOI: 10.3389/fnhum.2020.560021.
[27]	Elliott B L, D'Ardenne K, Mukherjee P, et al. Limbic and executive meso- and nigrostriatal tracts predict impulsivity differences in attention-deficit/hyperactivity disorder[J]. Biol Psychiatry Cogn Neurosci Neuroimaging, 2022, 7(4): 415-423. DOI: 10.1016/j.bpsc.2021.05.002.
[28]	李梦滔, 代梦琴, 罗尧竞, 等. 学龄期注意缺陷多动障碍儿童社交焦虑现状及影响因素[J]. 中国儿童保健杂志, 2024, 32(3): 344-348. DOI: 10.11852/zgetbjzz2023-0613.
[29]	Șipoș R, Văidean T, Răpciuc A M, et al. Analysing digital engagement patterns: a machine learning investigation into social anxiety among adolescents with ADHD[J]. J Clin Med, 2024, 13(23): 7461. DOI: 10.3390/jcm13237461.
[30]	Park C, Rouzi M D, Atique M M U, et al. Machine learning-based aggression detection in children with ADHD using sensor-based physical activity monitoring[J]. Sensors(Basel), 2023, 23(10): 4949. DOI: 10.3390/s23104949.
[31]	Sudre G, Sharp W, Kundzicz P, et al. Predicting the course of ADHD symptoms through the integration of childhood genomic, neural, and cognitive features[J]. Mol Psychiatry, 2021, 26(8): 4046-4054. DOI: 10.1038/s41380-020-00941-x.
[32]	Peterson B S, Trampush J, Maglione M, et al. Treatments for ADHD in children and adolescents: a systematic review[J]. Pediatrics, 2024, 153(4): e2024065787. DOI: 10.1542/peds.2024-065787.
[33]	Chekroud A M, Bondar J, Delgadillo J, et al. The promise of machine learning in predicting treatment outcomes in psychiatry[J]. World Psychiatry, 2021, 20(2): 154-170. DOI: 10.1002/wps.20882.
[34]	Feng A C, Zhi D M, Feng Y, et al. Functional imaging derived ADHD biotypes based on deep clustering: a study on personalized medication therapy guidance[J]. EClinicalMedicine, 2024, 77: 102876. DOI: 10.1016/j.eclinm.2024.102876.
[35]	苏怡, 王业鑫, 杨莉, 等. 临床和认知特征指标预测注意缺陷多动障碍药物疗效[J]. 中国心理卫生杂志, 2022, 36(3): 211-216.
[36]	Yoo J H, Sharma V, Kim J W, et al. Prediction of sleep side effects following methylphenidate treatment in ADHD youth[J]. Neuroimage Clin, 2020, 26: 102030. DOI: 10.1016/j.nicl.2019.102030.
[37]	白语殷, 冯敏妍, 毕波. 青少年精神疾病的跨诊断临床分期:国际专家共识(2020版)解读[J]. 中华实用儿科临床杂志, 2021, 36(22): 1716-1720.
[38]	Allesøe R L, Thompson W K, Bybjerg-Grauholm J, et al. Deep learning for cross-diagnostic prediction of mental disorder diagnosis and prognosis using Danish nationwide register and genetic data[J]. JAMA Psychiatry, 2023, 80(2): 146-155. DOI: 10.1001/jamapsychiatry.2022.4076.
[39]	Mcgorry P D, Nelson B. Transdiagnostic psychiatry: premature closure on a crucial pathway to clinical utility for psychiatric diagnosis[J]. World Psychiatry, 2019, 18(3): 359-360. DOI: 10.1002/wps.20679.
[40]	王月, 李赟, 王瑶, 等. 共情联合执行功能对高功能孤独症谱系障碍与注意缺陷多动障碍的诊断分类研究[J]. 中华行为医学与脑科学杂志, 2020(2): 120-124.
[41]	Schirmer M D, Venkataraman A, Rekik I, et al. Neuropsychiatric disease classification using functional connectomics - results of the connectomics in neuroimaging transfer learning challenge[J]. Med Image Anal, 2021, 70: 101972. DOI: 10.1016/j.media.2021.101972.
[42]	Ehrig L, Wagner A C, Wolter H, et al. FASDetect as a machine learning-based screening app for FASD in youth with ADHD[J]. NPJ Digit Med, 2023, 6(1): 130. DOI: 10.1038/s41746-023-00864-1.

领域	AI技术	第一作者	最优模型	性能
领域	AI技术	第一作者	最优模型	区分度	灵敏度	特异度	PPV	NPV	校准度	DCA	准确率
AI与可穿戴设备	ML	Kim^[8]	LightGBM	0.80	0.76	0.72	0.16	0.98	—	—	—
	DL	Muñoz-Organero^[9]	CNN（腕部）	≈0.80	0.60	1.00	—	—	—	—	0.88
			CNN（踝部）	≈0.90	0.80	1.00	—	—	—	—	0.94
AI与磁共振成像	ML	陈丽舟^[17]	SVM	0.87	0.89	0.71	—	—	—	—	0.76
	DL	朱莉^[15]	CNN	—	0.82	0.73	0.27	0.18	—	—	0.77
		Ke^[16]	Deep CSAF	≈1.00	1.00	—	1.00	—	—	—	1.00
AI与脑电图	ML	García-Ponsoda^[20]	XGBoost	—	—	—	—	—	—	—	0.86
	DL	Sanchis^[22]	EEG-MHCNet	—	0.855 1	—	0.781 3	—	—	—	0.760 9
	可解释XAI模型	Khare^[23]	EBM	1.00	0.997 8	0.998 4	0.998 7	0.13	—	—	0.998 1
AI与ADHD症状特征预测	ML	Elliott^[27]	LASSO回归模型	—	—	—	—	—	—	—	—
	DL	Park^[30]	随机森林模型	0.893	0.85	0.80	—	—	—	—	—
	ML-DL混合模型	Slobodin^[26]	随机森林+神经网络模型	—	0.89	0.84	—	—	—	—	0.87
AI与ADHD药物疗效预测	ML	Yoo^[36]	J48决策树	0.92	0.867	0.857	—	—	—	—	0.861
AI与ADHD药物疗效预测	DL	Feng^[34]	图卷积网络（GCN-BSD）	—	—	—	—	—	—	—	—
AI在ADHD跨诊断	ML	王月^[40]	SVM	—	—	—	—	—	—	—	0.627 5
		Schirmer^[41]	长短期记忆网络（LSTM）	—	—	—	—	—	—	—	—
		Ehrig^[42]	随机森林模型	0.92	0.00	—	0.86	—	0.11	—	0.85

领域	AI技术	第一作者	最优模型	性能
领域	AI技术	第一作者	最优模型	区分度	灵敏度	特异度	PPV	NPV	校准度	DCA	准确率
AI与可穿戴设备	ML	Kim^[8]	LightGBM	0.80	0.76	0.72	0.16	0.98	—	—	—
	DL	Muñoz-Organero^[9]	CNN（腕部）	≈0.80	0.60	1.00	—	—	—	—	0.88
			CNN（踝部）	≈0.90	0.80	1.00	—	—	—	—	0.94
AI与磁共振成像	ML	陈丽舟^[17]	SVM	0.87	0.89	0.71	—	—	—	—	0.76
	DL	朱莉^[15]	CNN	—	0.82	0.73	0.27	0.18	—	—	0.77
		Ke^[16]	Deep CSAF	≈1.00	1.00	—	1.00	—	—	—	1.00
AI与脑电图	ML	García-Ponsoda^[20]	XGBoost	—	—	—	—	—	—	—	0.86
	DL	Sanchis^[22]	EEG-MHCNet	—	0.855 1	—	0.781 3	—	—	—	0.760 9
	可解释XAI模型	Khare^[23]	EBM	1.00	0.997 8	0.998 4	0.998 7	0.13	—	—	0.998 1
AI与ADHD症状特征预测	ML	Elliott^[27]	LASSO回归模型	—	—	—	—	—	—	—	—
	DL	Park^[30]	随机森林模型	0.893	0.85	0.80	—	—	—	—	—
	ML-DL混合模型	Slobodin^[26]	随机森林+神经网络模型	—	0.89	0.84	—	—	—	—	0.87
AI与ADHD药物疗效预测	ML	Yoo^[36]	J48决策树	0.92	0.867	0.857	—	—	—	—	0.861
AI与ADHD药物疗效预测	DL	Feng^[34]	图卷积网络（GCN-BSD）	—	—	—	—	—	—	—	—
AI在ADHD跨诊断	ML	王月^[40]	SVM	—	—	—	—	—	—	—	0.627 5
		Schirmer^[41]	长短期记忆网络（LSTM）	—	—	—	—	—	—	—	—
		Ehrig^[42]	随机森林模型	0.92	0.00	—	0.86	—	0.11	—	0.85

人工智能在注意缺陷多动障碍儿童医学应用中的研究进展

Progress of Artificial Intelligence in Children with Attention Deficit Hyperactivity Disorder

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 1

参考文献 42

相关文章 15

编辑推荐

Metrics

留言

[1]	王振宇, 郭鹏飞, 郝文辉. 医学院校教师人工智能伦理素养研究[J]. 中国全科医学, 2026, 29(33): 4914-4921.
[2]	田向阳, 周迎生, 高莹. 糖尿病行为与生活方式干预展望[J]. 中国全科医学, 2026, 29(31): 4513-4520.
[3]	魏学娟, 王利华, 王丽, 任静林, 闫海青, 于海洋, 刘景峰. 智能语音外呼辅助传统干预对社区老年"三高"共病患者健康管理的效果研究：一项随机对照试验[J]. 中国全科医学, 2026, 29(31): 4635-4640.
[4]	谢远玲, 程丽, 龙湘玉, 苏政欣, 苏毅仪, 张人月, 孙国强, 陈旭霞, 周继兰. 人工智能眼科远程医疗诊断平台在社区眼健康筛查中的应用效果研究[J]. 中国全科医学, 2026, 29(31): 4641-4649.
[5]	黄牧, 张道春, 阳维, 钟丽明, 袁文静, 贾子琪, 谭相良, 段小慧, 刘岘, 陈维翠. 基于融合三维U-Net与Transformer的深度学习网络对腹部CT图像中结直肠癌病灶的自动分割：一项多中心、多设备研究[J]. 中国全科医学, 2026, 29(29): 4281-4287.
[6]	陈毓婧, 王岩, 李光, 董本超, 杨培川, 马剑雄. 数字技术在亚洲地区老年人肌少症干预中应用的范围综述[J]. 中国全科医学, 2026, 29(29): 4200-4210.
[7]	陈澄, 陈开元, 孙乃玲, 曾华堂, 雷苏文, 梁万年. 人工智能赋能疾病预防控制标准化建设现状及建议[J]. 中国全科医学, 2026, 29(29): 4193-4199.
[8]	邓玮, 黄梦婷, 易胜利, 林杨, 许力丹, 任哲学, 朱秀琼, 李烨莎, 任梦鸽, 柯大智, 李桂琼, 吴志勤, 骆傲然, 何发培. AI赋能全科临床思维的培养与实施效果研究[J]. 中国全科医学, 2026, 29(28): 4078-4085.
[9]	解霄芸, 刘晨曦, 柳羽姝, 周艺涵. 中国基层医疗卫生机构人工智能研究实践及潜在数据贫困问题：一项范围综述[J]. 中国全科医学, 2026, 29(28): 4067-4077.
[10]	姜妍琳, 王喆, 翟睿, 张梦洁, 陈溪, 王俊宏. 中药复方治疗注意缺陷多动障碍优势方案筛选研究：基于网状Meta分析和熵权TOPSIS法[J]. 中国全科医学, 2026, 29(27): 3985-3994.
[11]	周雨涵, 闵钰琦, 郭洲, 杨敏, 岳露, 何美忱, 周海蓉, 王永晨. 人工智能时代全科医生的机遇与挑战[J]. 中国全科医学, 2026, 29(25): 3563-3568.
[12]	王洪川, 张剑波, 马伟, 赵司迪. 数字政策引领医疗数据空间的治理逻辑与生态构建：基于"技术-管理-法律"的分析框架[J]. 中国全科医学, 2026, 29(24): 3416-3422.
[13]	欧阳微娜, 王露, 张璇, 唐薇, 范咏梅. 基于动态多尺度特征融合与注意力机制的心房颤动筛查研究[J]. 中国全科医学, 2026, 29(23): 3281-3286.
[14]	熊丹, 刘圆, 刘思敏, 吴建炜, 胡笑霞. 新媒体遗书中挖掘年轻人存在感空虚及其应对行为：人工智能辅助的内容分析[J]. 中国全科医学, 2026, 29(22): 3182-3191.
[15]	陈开元, 王洪川, 葛袆, 孙乃玲, 陈澄, 王天然, 曾华堂, 雷苏文, 梁万年. 生成式人工智能赋能卫生健康标准体系构建现状与挑战[J]. 中国全科医学, 2026, 29(22): 3073-3078.