人工智能技术在特定性学习障碍诊疗中的应用

Application of artificial intelligence in the diagnosis and treatment of specific learning disorder

摘要

特定性学习障碍(SLD)是学龄儿童中患病率最高的神经发育性障碍之一,其传统诊断方法存在耗时长、成本高、主观性强等局限。人工智能(AI)技术为SLD的早期识别、精准诊断与个性化干预提供了新途径。本文概述了AI在SLD诊疗中的主要应用:在辅助评估方面,结合眼动追踪技术可实现高生态效度的阅读障碍筛查,结合手写分析可量化静态字形特征与动态书写过程,结合语音声学特征分析可评估阅读流利性,结合神经影像学技术可揭示SLD的神经生物学特征;在干预方面,构建自适应智能辅导系统可实现个性化干预,利用大语言模型可为社会适应提供辅助支持。但目前仍面临数据集规模偏小、算法可解释性不足、伦理隐私风险及临床验证不充分等挑战。未来需构建大规模多中心中文数据集、发展低成本筛查工具、加强循证评价并完善监管框架,推动AI成为SLD诊疗体系变革的核心动力。

Abstract

Specific learning disorder (SLD) is one of the most prevalent neurodevelopmental disorders in school-aged children. However, conventional diagnostic workflows are frequently constrained by prolonged assessment cycles, high costs, and inherent subjectivity. Artificial intelligence (AI) has provided novel paradigms for the early identification, precision diagnosis, and personalized intervention of SLD. This review outlines the main applications of AI in the diagnosis and treatment of SLD. Regarding assisted assessment, the integration of eye-tracking technology facilitates dyslexia screening with enhanced ecological validity, while computerized handwriting analysis quantifies both static orthographic features and dynamic kinematic processes. Furthermore, acoustic analysis of speech patterns enables the objective evaluation of reading fluency, and advanced neuroimaging techniques can elucidate the underlying neurobiological characteristics of SLD. In terms of intervention, adaptive intelligent tutoring systems provide personalized intervention, and large language models offer promising auxiliary tools for social adaptation. Despite these advancements, significant challenges persist, including limited dataset scales, insufficient algorithmic interpretability, ethical concerns regarding data privacy, and inadequate clinical validation. Future efforts should prioritize the construction of large-scale, multi-center Chinese-specific datasets, the development of cost-effective screening tools, and the establishment of robust evidence-based evaluation and regulatory frameworks to promote AI as the core driving force for the transformation of the SLD diagnosis and treatment system.

关键词

特定性学习障碍 / 人工智能 / 早期筛查 / 个性化干预

Key words

specific learning disorder / artificial intelligence / early screening / personalized intervention

[1] Peters L, Op de Beeck H, De Smedt B. Cognitive correlates of dyslexia, dyscalculia and comorbid dyslexia/dyscalculia: Effects of numerical magnitude processing and phonological processing[J]. Res Dev Disabil, 2020, 107: 103806.
[2] McGrath LM, Stoodley CJ. Are there shared neural correlates between dyslexia and ADHD? A Meta-analysis of voxel-based morphometry studies[J]. J Neurodev Disord, 2019, 11(1): 31.
[3] 刘钟泠, 刘铎, 万勤, 等. 中外发展性阅读障碍评估方法对比[J]. 教育生物学杂志, 2023, 11(3): 236-242.
Liu ZL, Liu D, Wan Q, et al. Comparison of Chinese and foreign assessment methods for developmental dyslexia[J]. J Bio Educ, 2023, 11(3): 236-242. (in Chinese)
[4] Wang Y, Long J, Wang P. The prevalence of mathematical difficulties among primary school children in Mainland China: A systematic review and meta-analysis[J]. Front Public Health, 2024, 11: 1250337.
[5] Fletcher JM, Miciak J. Assessment of specific learning disabilities and intellectual disabilities[J]. Assessment, 2024, 31(1): 53-74.
[6] Hajovsky DB, Maki KE, Niileksela CR, et al. Science and practice of identifying specific learning disabilities: Kind conversations about a wicked problem[J]. J Psychoeduc Assess, 2025, 43(4): 359-379.
[7] Sreekumar S, Lijiya A, Ravindren RK. Bridging the gap: Cognitive science perspectives and artificial intelligence for prediction and detection of specific learning disabilities[J]. IEEE Trans Cogn Dev Syst, 2025, 17(4): 711-726.
[8] Shahini A, Kamath AP, Sharma E, et al. A systematic review for artificial intelligence-driven assistive technologies to support children with neurodevelopmental disorders[J]. Inf Fusion, 2025, 124: 103441.
[9] Ke F, Liu R, Sokolikj Z, et al. Using eye-tracking in education: Review of empirical research and technology[J]. Educ Technol Res Dev, 2024, 72(3): 1383-1418.
[10] Toki EI. Using eye-tracking to assess dyslexia: A systematic review of emerging evidence[J]. Educ Sci, 2024, 14(11): 1256.
[11] Ward LM, Kapoula Z. Dyslexics' fragile oculomotor control is further destabilized by increased text difficulty[J]. Brain Sci, 2021, 11(8): 990.
[12] Bonifacci P, Tobia V, Sansavini A, et al. Eye-movements in a text reading task: A comparison of preterm children, children with dyslexia and typical readers[J]. Brain Sci, 2023, 13(3): 425.
[13] Vajs I, Papic' T, Kovic' V, et al. Accessible dyslexia detection with real-time reading feedback through robust interpretable eye-tracking features[J]. Brain Sci, 2023, 13(3): 405.
[14] Shi W, Jiang X. Predicting Chinese reading proficiency based on eye movement features and machine learning[J]. Read Writ, 2025, 38(5): 1383-1407.
[15] Ramchandar R, Sanyuktha S, Mittal S, et al. Eye-tracking driven dyslexia detection: A data-efficient approach using synthetic augmentation and XGBoost classifier[C]//IEEE 25th International Conference on Bioinformatics and Bioengineering (BIBE). Athens, Greece. 2025: 81-85.
[16] Nerušil B, Polec J, Škunda J, et al. Eye tracking based dyslexia detection using a holistic approach[J]. Sci Rep, 2021, 11(1):15687.
[17] Svaricek R, Dostalova N, Sedmidubsky J, et al. INSIGHT: Combining fixation visualisations and residual neural networks for dyslexia classification from eye-tracking data[J]. Dyslexia, 2025, 31(1): e1801.
[18] Li X, Li Z, Xu F. Eye-tracking based detection of developmental dyslexia in children using convolutional-transformer network[C]//27th International Conference on Computer Supported Cooperative Work in Design (CSCWD). Tianjin, China. 2024: 821-826.
[19] Robaa M, Balat M, Awaad R, et al. Explainable AI in handwriting detection for dyslexia using transfer learning[C]//12th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC). Alexandria, Egypt. 2024: 17-22.
[20] Rangasrinivasan S, Sumi Suresh MS, Olszewski A, et al. AI-enhanced child handwriting analysis: A framework for the early screening of dyslexia and dysgraphia[J]. SN Comput Sci, 2025, 6(5).doi:https://doi.org/10.1007/s42979-025-03927-0.
[21] Xu Z, Mittal PS, Ahmed MM, et al. Assessing penmanship of Chinese handwriting: A deep learning-based approach[J]. Read Writ, 2025, 38(3): 723-743.
[22] Li-Tsang CWP, Li TMH, Yang CN, et al. Evaluating Chinese handwriting performance of primary school students using the Smart Handwriting Analysis and Recognition Platform (SHARP) [J]. Ann Pediatr and Child Health, 2022,10(5):1283.
[23] Rosenblum S, Dror G. Identifying developmental dysgraphia characteristics utilizing handwriting classification methods[J]. IEEE Trans Hum Mach Syst, 2017, 47(2): 293-298.
[24] Wong LH, Guat Poh AW, Sun H, et al. Effects of a machine learning-empowered Chinese character handwriting learning tool on rectifying legible writing in young children: A pilot study[J]. Int Conf Comput Educ, 2023,1:742-747.
[25] Serpa-Andrade LJ, Pazos-Arias JJ, Gil-Solla A, et al. An automatic feedback educational platform: Assessment and therapeutic intervention for writing learning in children with special educational needs[J]. Expert Syst Appl, 2024, 249: 123641.
[26] Sabu K, Rao P. Predicting children's perceived reading proficiency with prosody modeling[J]. Comput Speech Lang, 2024, 84: 101557.
[27] Chung WL, Bidelman GM. Acoustic features of oral reading prosody and the relation with reading fluency and reading comprehension in Taiwanese children[J]. J Speech Lang Hear Res, 2022, 65(1): 334-343.
[28] Miller J, Schwanenflugel PJ. A longitudinal study of the development of reading prosody as a dimension of oral reading fluency in early elementary school children[J]. Read Res Q, 2008, 43(4): 336-354.
[29] Chung WL. Oral reading prosody and the relation with reading abilities: A comparison of two rating scales[J]. Read Writ, 2024, 37(4): 869-886.
[30] Vitthal C, Shreeharsha SB, Sabu K, et al. Predicting lexical skills from oral reading with acoustic measures[EB/OL].(2021-12-01)[2026-04-14]https://arxiv.org/abs/2112.00635.
[31] da Silva GC, Rodrigues RL, Amorim AN, et al. Assessing reading fluency in elementary grades: A machine learning approach[J]. Comput Educ Artif Intell, 2025, 8: 100411.
[32] Phatangare S, Jain S, Chakrabarty S, et al. LexiEase AI: An AI powered dyslexic support system[C]//5th International Conference on Trends in Material Science and Inventive Materials (ICTMIM).Kanyakumari, India.2025: 1412-1417.
[33] Zhai J, Zhou Z, Zhao L. Application of intelligent speech recognition technology based on wavelet algorithm in Chinese oral teaching[C]//International Conference on Electrical Drives, Power Electronics & Engineering (EDPEE). Athens, Greece. 2024: 976-981.
[34] Martínez-Briones BJ, Fernández T, Silva-Pereyra J. Electroencephalographic power spectrum patterns related to the intelligence of children with learning disorders[J]. Peer J, 2025, 13: e19138.
[35] Rodríguez-Rodríguez I, Mateo-Trujillo JI, Ortiz A, et al. Directed weighted EEG connectogram insights of one-to-one causality for identifying developmental dyslexia[J]. Int J Neur Syst, 2025, 35(6): 2550032.
[36] Usman OL, Muniyandi RC, Omar K, et al. Advance machine learning methods for dyslexia biomarker detection: A review of implementation details and challenges[J]. IEEE Access, 2021(9): 36879-36897.
[37] Christodoulides P, Miltiadous A, Tzimourta KD, et al. Classification of EEG signals from young adults with dyslexia combining a brain computer interface device and an interactive linguistic software tool[J]. Biomed Signal Process Control, 2022,76: 103646.
[38] Guhan Seshadri NP, Agrawal S, Kumar Singh B, et al. EEG based classification of children with learning disabilities using shallow and deep neural network[J]. Biomed Signal Process Control, 2023,82: 104553.
[39] Prasad S, Sagar R, Kumaran SS, et al. Study of functional magnetic resonance imaging (fMRI) in children and adolescents with specific learning disorder (dyslexia)[J]. Asian J Psychiatry, 2020,50: 101945.
[40] Zahia S, Garcia-Zapirain B, Saralegui I, et al. Dyslexia detection using 3D convolutional neural networks and functional magnetic resonance imaging[J]. Comput Meth Programs Biomed, 2020,197: 105726.
[41] Khan OA, Peters L, Kwok FY, et al. Neural connectivity in developmental dyscalculia: A resting-state fMRI study using ROI-based and connectome-based approaches[J]. ZDM Mathematics Education, 2026, 58(3): 399-413.
[42] Tomaz Da Silva L, Esper NB, Ruiz DD, et al. Visual explanation for identification of the brain bases for dyslexia on fMRI data[J/OL].Front Comput Neurosci,2021,15:594659.doi: 10.3389/fncom.2021.594659.
[43] Alkhurayyif Y, Wahab Sait AR. Deep learning-driven dyslexia detection model using multi-modality data[J]. Peer J Computer Science, 2024, 10: e2077.
[44] Wilson AJ, Dehaene S, Pinel P, et al. Principles underlying the design of “The Number Race”, an adaptive computer game for remediation of dyscalculia[J]. Behav Brain Funct, 2006,2: 19.
[45] Fan M, Wu Y, Fan J, et al. Character alive: Designing and evaluating a tangible system to support children's Chinese radical and character learning[J]. Int J Hum, 2025, 41(2): 1293-1308.
[46] Liu L, Fang T, Liu E, et al. CNReader: A reading practice tool designed for Chinese children with developmental dyslexia[J].Humanit Soc Sci Commun, 2025,12(1):751.
[47] Ahuja NJ, Dutt S, Choudhary SL, et al. Intelligent tutoring system in education for disabled learners using human-computer interaction and augmented reality[J].International Journal of Human-Computer Interaction, 2025, 41(3): 1804-1816.
[48] Paglialunga A, Melogno S. The effectiveness of artificial intelligence-based interventions for students with learning disabilities: A systematic review[J]. Brain Sci, 2025, 15(8): 806.
[49] Goodman SM, Buehler E, Clary P, et al. LaMPost: AI writing assistance for adults with dyslexia using large language models[J]. Commun ACM, 2024, 67(9): 80-89.
[50] Yan K, Chiu K, Hang L, et al. A study on using ChatGPT to help students with dyslexia learn Chinese and English writing[J].ECNU Rev Educ, 2026(9): 20965311251358269.
[51] Egbuna IK, Otoibhili PA, Abdulkareem RO, et al. Explainable artificial intelligence (XAI) in diagnosing neurodevelopmental disorders: From black boxes to clinical transparency[J]. Int J Biol Pharm Sci Arch, 2025, 10(1): 31-57.
[52] Lekadir K, Feragen A, Fofanah AJ, et al. FUTURE-AI: International consensus guideline for trustworthy and deployable artificial intelligence in healthcare[J]. BMJ,2023(388):e081554.

基金资助

上海市东方英才计划拔尖项目(2023-2-26);上海申康医院发展中心市级医院新兴前沿技术联合攻关项目(SHDC12026115);上海交通大学“交大之星”计划医工交叉中心项目(2024-2)