目的 基于机器学习探索数字划消测验用于学习障碍(LD)快速筛查的可行性研究,为探索LD的快速筛查提供新的研究思路和工具。方法 利用数字划消测验,2020年9月—2021年3月对南京市区随机选取的414名7~12岁儿童进行调查,首先获得数字划消测验的18个观察指标;将这些观察指标作为特征,探索利用7种经典的机器学习算法(包括线性判别分析、朴素贝叶斯、K近邻、神经网络、支持向量机、决策树和随机森林),实现对学习障碍3种亚型的分类;比较不同机器学习算法的分类性能,以从中选出推荐算法。结果 在7种经典的机器学习算法中,随机森林的优势突出,其准确率和曲线下面积(AUC)值分别达0.83和0.92。结论 借助随机森林算法,利用数字划消测验(只需2min就可完成)可以实现对学习障碍的快速筛查,进一步证实机器学习应用于疾病预测的可行性。
Abstract
Objective To explore the feasibility of a rapid screening method using number cancellation test (NCT) in combination with machine learning algorithm for learning disabilities (LD), in order to provide ideas for rapid screening of LD. Methods A total of 414 children aged 7-12 years were randomly recruited in Nanjing, and were asked to complete the NCT, in which 18 parameters were measured to evaluate individual's abilities during NCT. Then, these parameters were considered as classification features to explore the classification of LD by seven machine learning algorithms, including linear discriminant analysis, naive Bayes, K-nearest neighbor, neural network, support vector machine, decision tree and random forest. Finally, the classification performance among different machine learning algorithms was compared. Result Among the seven machine learning algorithms, random forest was outstanding for screening LD, and the accuracy and AUC were 0.83 and 0.92, respectively. Conclusion NCT in combination with random forest algorithm can be applied as a rapid screening method (completed within 2 minutes) for LD,which further proves the feasibility of machine learing in disease prediction.
关键词
学习障碍 /
机器学习 /
快速筛查 /
数字划消 /
学龄儿童
Key words
learning disabilities /
machine learning /
rapid screening /
number cancellation test /
school-age children
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 杨斌让. 特定学习障碍[J]. 中华实用儿科临床杂志,2015,30(11):810-817.
Yang BR. Specific learning disorder[J]. Chin J Appl Clin Pediatr, 2015,30(11):810-817.(in Chinese)
[2] Miciak JJM, Fletcher JM. The critical role of instructional response for identifying dyslexia and other learning disabilities[J]. J Learn Disabil, 2020,53(5):343-353.
[3] 静进,海燕,邓桂芬,等. 学习障碍筛查量表的修订与评价[J]. 中华儿童保健杂志,1998,6(3):197-200.
Jing J, Hai Y, Deng GF, et al. The revision and appraisal of thePupil Rating Scale revised-screening for learning disabilities[J]. Chin J Child Health Care,1998,6(3):197-200.(in Chinese)
[4] 王忠,静进,艾素英,等. 学习障碍儿童筛查量表区域性实施的信度与效度分析[J]. 中国预防医学杂志,2010,11(7):682-685.
Wang Z, Jing J, Ai SY, et al. Reliability and validity of Revised Pupil Rating Scale in regional implementation of learning disability screening[J]. Chin Prev Med,2010,11(7):682-685. (in Chinese)
[5] Xie Y, Wang H, Chen Y, Liu F, et al. Establishing normative data for the number cancellation test among children in kindergarten and primary schools in China [J]. Front Psychiatry, 2022,13:788825.
[6] 陈雨欣,王洪安,刘福临,等. 54名注意缺陷多动障碍儿童数字划消测试的时空特征分析[J].教育生物学, 2022,10(3):182-187.
Chen YX, Wang HA, Liu FL, et al. Analysis of spatiotemporal features in number cancellation test for 54 children with attention-deficit/hyperactivity disorder[J]. Journal of Bio-education,2022,10(3):182-187.(in Chinese)
[7] Gupta C, Chandrashekar P, JinT, et al. Bringing machine learning to research on intellectual and developmental disabilities:Taking inspiration from neurological diseases[J]. J Neurodev Disord, 2022, 14(1):1-22.
[8] Weeks S, Atlas A. Clinical audit for occupational therapy intervention for children with autism spectrum disorder:Sampling steps and sample size calculation [J]. BMC Res Notes, 2015, 8:282.
[9] 刘姿镃,徐小雨,潘宁,等. 儿童认知能力与学习障碍的关联性研究[J]. 中国儿童保健杂志, 2022,30(8):834-838.
Liu ZZ, Xu XY, Pan NN, et al. Study on the correlation between cognitive ability and learning disability in children[J]. Chin J Child Health Care, 2022,30(8):834-838.(in Chinese)
[10] Kirtania R, Mitra S, Shankar BU. A novel adaptive k-NN classifier for handling imbalance:Application to brain MRI[J]. Intell Data Anal, 2020, 24(4):909-924.
[11] Sachser C, Berliner L, Risch E, et al. The child and Adolescent Trauma Screen 2 (CATS-2) - validation of an instrument to measure DSM-5 and ICD-11 PTSD and complex PTSD in children and adolescents[J]. Eur J Psychotraumatol, 2022, 13(2):2105580.
[12] 王彦光, 朱鸿斌, 徐维超. ROC曲线及其分析方法综述 [J]. 广东工业大学学报, 2021, 38(1):46-53.
Wang YG, Zhu HB, Xu WC. A review on ROC curve and analysis[J]. Journal of Guangdong University of Technology,2021,38(1):46-53.(in Chinese)
[13] 吴军, 欧阳艾嘉, 张琳. 基于标准置换检验的差异序列模式挖掘算法 [J]. 计算机应用研究, 2021,38(3):710-713.
Wu J, Ouyang AJ, Zhang L. Mining discriminative sequential patterns based on standard permulation testing[J]. Application Research of Computers,2021,38(3):710-713.(in Chinese)
[14] Fisher PW, Reyes-Portillo JA, Riddle MA,et al. Systematic review:Nonverbal learning disability [J]. J Am Acad Child Psy, 2021, 61(2):159-186.
[15] 贾芷莹,董旻晔,施贞夙,等. 基于机器学习的轻度认知功能障碍筛查研究[J]. 上海交通大学学报(医学版),2019,39(8):908-913.
Jia ZY, Dong MY, Shi ZS, et al. Study of a screening system for mild cognitive impairment based on machine learning model[J]. Journal of Shanghai Jiao Tong University (Medical Science),2019,39(8):908-913.(in Chinese)
[16] 张家宁,张静,蒋雅丽,等. 机器学习在注意缺陷多动障碍与自闭症谱系障碍的辅助诊断[J]. 中华行为医学与脑科学杂志,2017,26(8):754-759.
Zhang JN, Zhang J, Jiang YL, et al. The computer-aided diagnosis of attention-deficit hyperactivitydisorder and autism spectrum disorder based on structural magnetic resonance imaging[J]. Chin J Behav Med Brain Sci,2017,26(8):754-759. (in Chinese)
[17] 刘笑晗,陈明隆,郭静. 机器学习在儿童创伤后应激障碍识别及转归预测中的应用[J]. 心理科学进展,2022,30(4):851-862.
Liu XH, Chen ML, Guo J. Application of machine learning in prognosis and trajectory of post-traumatic stress disorder in children[J]. Advances in Psychological Science,2022,30(4):851-862.(in Chinese)
[18] 范炤,李彩.基于机器学习的阿尔茨海默病病程分类[J]. 中国医学影像学杂志, 2019,27(10):792-795,800.
Fan Z, Li C. Classification of Alzheimer's disease course based on machine learning[J]. Chinese Journal of Medical Imaging,2019,27(10):792-795,800.(in Chinese)
基金
国家自然科学基金(62073077、61673113)
PDF(713 KB)
