机器学习在发展性阅读障碍儿童早期筛查中的应用  

Application of machine learning in early screening of children with dyslexia

作　　者：卜晓鸥 王耀 杜亚雯 王沛 BU Xiaoou;WANG Yao;DU Yawen;WANG Pei(Department of Special Education,Faculty of Education,East China Normal University,Shanghai 200062,China)

机构地区：[1]华东师范大学教育学部特殊教育学系,上海200062

出　　处：《心理科学进展》2023年第11期2092-2105,共14页Advances in Psychological Science

基　　金：“华东师范大学幸福之花‘音乐画的脑智机制及促进儿童艺术教育发展的实践进路’”资助。

摘　　要：发展性阅读障碍严重影响儿童的学业成就、心理健康和社会适应能力。近年来,机器学习因其强大的数据处理和挖掘能力逐渐被应用到阅读障碍儿童的早期筛查中,在标准化心理教育测试、眼动追踪、游戏测试、脑成像等多个领域积累了较为丰富的成果,获得了更加精准高效、灵活可靠的分类结果。然而,机器学习在对象选取、数据采集、转化潜力和安全隐私等方面仍存在局限性。未来研究需要重点关注学龄前阅读障碍儿童的早期筛查系统的科学性,同时积极构建多模态数据库、在多种算法中寻找最佳算法以获取最优参数,最终实现临床实践中的广泛使用。Developmental dyslexia is the most prevalent form of specific learning disorder with a neurobiological basis that not only restricts an individual's academic achievement and career development,but also negatively affects an individual's psychological and social adjustment substantially.Recently,machine learning has been gradually applied to the early screening of children with dyslexia due to its powerful data processing and mining capabilities,accumulating richer results in various aspects such as standardized psychoeducational testing,eye tracking,game testing and brain imaging.However,machine learning still has limitations in terms of participant selection,data collection,transformation potential,security and privacy.Future researches need to focus on the early identification of pre-school children with dyslexia,construct multimodal data,and find the best classifier among multiple classifiers to obtain optimal parameters,which will eventually achieve widespread use in clinical practice.

关 键 词：发展性阅读障碍 机器学习 早期筛查 儿童 

分 类 号：R395[哲学宗教—心理学]